The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
This role is responsible for advancing the agentic capabilities of our models, with ownership spanning the full development cycle. Our research team is small, and the role carries a corresponding degree of autonomy and responsibility.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest in this research area. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Build a synthetic data framework used across the team, and create new task environments to scale training across agentic capabilities such as tool use, long-horizon tasks, and complex workflows.
Address identified gaps through data and recipe work, validating proposed changes through controlled ablations.
Enhance usability for agent capabilities end to end, translating internal and external feedback into model improvements.
Minimum qualifications:
Strong engineering skills, ability to contribute code and debug in complex codebases.
Ability to design, run, and interpret experiments with scientific rigor and clarity.
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
Experience building synthetic data pipelines and systems that were adopted by others on your team and remain in use today.
Experience owning the end-to-end cycle of identifying gaps in model usability and closing them through custom evaluations and training data.
Experience making large-scale agentic RL infrastructure reliable given the long tail of failures that surface at scale.
Experience improving the agentic capabilities of a frontier model.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're hiring a Site Reliability Engineer (SRE) to keep our post-training and reinforcement learning (RL) systems fast, reliable, and easy for researchers to iterate on. Think of this as a production engineering or site reliability role built around model training: you'll own the health of the training runs, clusters, and pipelines that power post-training and RL at Thinking Machines.
You'll work side by side with research teams during active model runs — debugging failures in real time, hardening infrastructure against the next class of problem, and building the tooling and automation that let researchers spend their time on the science instead of babysitting jobs. This role has real ownership: you'll be the person a research team calls when a run stalls at 2am, and the person who makes sure it doesn't happen again.
Own the reliability, performance, and uptime of large-scale post-training and RL training jobs, from launch through completion
Partner directly with research teams during active model runs, embedding with them to unblock training and speed up iteration
Debug failures across the full stack — accelerators, networking, storage, schedulers, and training frameworks — and drive issues to root cause
Build monitoring, alerting, and automated recovery so runs self-heal or fail fast instead of silently stalling
Improve checkpointing, fault tolerance, and job scheduling so hardware failures cost minutes, not days of compute
Build internal tools that reduce toil and improve cluster utilization across post-training and RL workloads
Participate in an on-call rotation supporting production model runs
Write postmortems and turn recurring failure patterns into permanent infrastructure fixes
4+ years of experience as a production engineer, site reliability engineer, or infrastructure engineer operating large-scale distributed systems in production
Track record debugging complex failures across distributed systems — networking, hardware, kernel, or scheduler issues
Strong software engineering skills in Python and/or Go/C++, with the judgment to know when to script a fix versus build a system
Solid grounding in Linux systems internals and networking fundamentals
Comfortable owning production systems, including participating in on-call rotations
Experience operating GPU or TPU training clusters at scale
Familiarity with post-training and RL techniques (e.g., RLHF, PPO, DPO) and the infrastructure challenges specific to them, such as reward model serving, rollout generation, and mixed training/inference workloads
Experience with distributed training frameworks (e.g., PyTorch, Ray) and job schedulers (e.g., Slurm, Kubernetes)
Experience with high-performance networking (e.g., InfiniBand, RDMA, NCCL) and its role in distributed training performance
Experience building observability tooling purpose-built for ML training, not just general infrastructure
A track record of thriving in fast-changing, research-driven environments where priorities shift with the science
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're looking for a network engineer to own the lowest layers of the network stack that our large-scale training and inference depend on. A single degraded link or flapping NIC can quietly slow a long training run or take it down outright; you'll be responsible for interconnect reliability at scale, across large GPU fabrics — both the RDMA/RoCE fabric between nodes and the NVLink/NVSwitch domains within them.
This is a hands-on, cross-stack role. You'll debug production collectives down to the NIC, build the instrumentation and tooling that makes the next debugging session dramatically faster, and serve as the technical point of contact who drives issues to resolution with our cloud providers' networking teams. Your goal is for our researchers to trust the fleet without worrying about the fabric underneath.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Reason about and validate GPU network fabric design across our deployments.
Debug RDMA / RoCEv2 across different NIC vendors. Diagnose collective failures of production NCCL, PFC/ECN tuning, and congestion control behavior.
Own NVLink / NVSwitch interconnect — including fabric manager and IMEX health, link and lane errors, and how the GPU fabric interacts with collectives.
Build host-level network instrumentation and use Linux tooling to build dashboards and alerts, not just the bug report.
Navigate cross-cloud fabric quirks across providers and triage across the NIC, driver, kernel, switch, and workload boundaries.
Drive escalations with cloud-provider networking teams, owning issues end-to-end until they're resolved.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Experience operating large‑scale clusters and container orchestration systems (e.g. Kubernetes or Slurm).
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Fluency with host-level debugging tools on Linux.
Strong communication skills, internally and with cloud providers.
Extensive experience with at least one of the following:
Familiarity with cloud network primitives across at least two cloud providers.
Hands-on experience with NVLink / NVSwitch, fabric manager, and IMEX.
Statistical rigor in reliability reasoning — comfort reasoning about failure and error rates, distributions, and base rates, and the judgment to separate signal from noise when characterizing a large fabric.
A track record of writing tooling that made the next debugging session meaningfully faster.
Familiarity with CUDA/NCCL and performance profiling for distributed training and inference.
Understanding of deep learning frameworks and their underlying system architectures.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're hiring an engineer to ensure the reliability of our GPU supercomputing fleet, owning the seam between hardware, firmware, and operating system. You will track the long tail of hardware issues: We are conducting frontier research in AI and a single bad NIC, HBM or a kernel driver edge case can compromise an experiment. Your job is to diagnose these issues, track their root cause down to the hardware, and resolve them internally or directly with vendors so that our researchers can run at scale and with confidence.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Investigate, reproduce, and remediate issues across large GPU clusters.
Own the drivers, kernel surface, and diagnostics that span hardware, firmware, and OS.
Automate the monitoring of fleet reliability and analyze error rates to validate whether a fix or firmware change measurably reduced failures rather than shifting them around.
Drive the firmware lifecycle: tracking, qualification, staged rollout, and regression analysis.
Engage vendors directly — GPUs, server OEMs, NIC vendors, and storage vendors — to get real fixes rather than ticket numbers. Manage RMA flows when hardware needs to come out.
Monitor and improve GPU hardware health signals and turn them into actionable reliability improvements.
Write clear postmortems and vendor cases that move issues forward.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Experience operating large‑scale clusters and container orchestration systems (e.g. Kubernetes or Slurm).
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Fluency with Linux systems and debugging tools.
Proven statistical rigor in analyzing reliability.
A track record of debugging a problem from application symptom to the root cause in hardware.
Comfort reading vendor errata, firmware release notes, and kernel changelogs.
Experience engaging hardware vendors directly — not just through escalation portals.
Linux kernel literacy: the scheduler, memory management, IRQ paths, and the driver model.
Out-of-band management experience: BMC / iDRAC / IPMI / Redfish.
Depth in GPU hardware health: Xid error taxonomy, NVLink, NVSwitch, fabric manager, and DCGM.
Proficiency in at least one backend language (we use Python and Rust).
Significant ownership of the hardware reliability function at scale.
Strong writing skills for vendor cases and postmortems.
An instinct for telling apart a flaky machine, a flaky workload, and a flaky test.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for a Safety Operations Lead who will focus on making our products safe by default while supporting fast product iteration and ambitious ideas.
You’ll work closely with product engineers, security, researchers, and designers to bake safety and integrity into the way we design, build, and ship human-AI collaboration tools. Day to day, you'll be in the moderation queue while building the tooling and policy that make the next round of moderation faster and more accurate.
Review flagged content, safety escalations, and account-level abuse signals daily. This is a standing responsibility, not a rotation: you'll triage cases, apply policy judgment, and take action (content flags, account review, bans/recovery) on an ongoing basis.
Use patterns from that casework to design and refine safety policy across the product stack, working with engineering, legal, safety research, and security stakeholders.
Build and maintain tooling and automation that make your own casework faster and more consistent: triage agents, ban/recovery workflows, abuse detection frameworks and templates.
Partner with product teams to embed safety into the product experience: model refusals, content flagging, account review, and safety protections, informed by what you're seeing in the queue.
Improve observability and detection for safety-relevant events (model safety trends, abuse patterns, malicious behavior in production), so the next round of cases surfaces faster and with better signal.
Minimum qualifications:
7+ years in an operational trust & safety, content moderation, or fraud/abuse ops role with direct, recurring responsibility for a case queue.
Experience owning policy definition, operationalization, and enforcement end to end, evidenced by specific policies or enforcement programs you've built or run.
Direct case experience with at least one of: cybersecurity abuse, CBRN-relevant risk, youth safety, or prompt injection, in a production environment.
Working familiarity with model safety and abuse risk categories (jailbreaks, prompt injection, scaled abuse) and how to identify and mitigate them in a live product, evidenced by specific cases you've handled.
Practical experience using AI tools (Claude, Codex, or similar) to build or accelerate operational workflows, not just as a general user of these tools.
Preferred qualifications:
We encourage you to apply even if you don’t meet all preferred qualifications.
Experience with safety and integrity operations specifically on AI-powered products or LLM APIs, and their unique abuse patterns.
Track record of turning recurring case patterns into reusable tooling, workflows, or process improvements, while still owning the underlying queue rather than handing it off once the interesting part is solved.
Experience training, onboarding, or setting the quality bar for other moderators or reviewers, showing you scale a team's output rather than just your own.
You have thoughtful opinions about what safe, trustworthy, frictionless user experiences look like, and you test those opinions against real cases in the queue, not just in the abstract.
You can translate safety and technical constraints into clear product trade-offs and feature requirements, but you see that as something that grows out of daily casework, not a substitute for it.
You bias toward speed and learning, and you measure that bias by how much faster or better the queue runs this month, not by how many new risk categories you've personally discovered.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $190,000 - $300,000.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for generalist infrastructure and systems engineers to help build the systems that power our foundation models and the internal teams on research and product development to be able to create the models and ship the products powered by our models.
You'll join a small, high-impact team responsible for architecting and scaling the core infrastructure behind everything we do. You’ll work across the full technical stack, solving complex distributed systems problems and building robust, scalable platforms.
Infrastructure is critical to us: it's the bedrock that enables every breakthrough. You'll work directly with researchers to accelerate experiments, improve infrastructure efficiency, and enable key insights across our models, products, and data assets.
We interview generally, but during project selection we’ll take into account your interests and experience alongside organizational needs. This flexible approach allows us to match talented engineers with the infrastructure teams where they'll have the greatest impact and growth potential.
Here are example areas you may contribute to depending on your area of expertise and interest:
Core Infrastructure: We support teams that train, research, and ultimately serve AI models and build the underlying infrastructure for the clusters to reliably and safely train frontier models. Examples might include building systems and running large Kubernetes clusters with GPU workloads, or building infrastructure to support Tinker.
Data Infrastructure: We build and maintain the data systems for our research and products. You'll design and optimize data pipelines using tools like Spark and other modern data infrastructure technologies. You’ll build scalable, reliable, data infrastructure while embedding governance best practices.
Developer Productivity: We care deeply about research and engineering productivity and our ability to continue shipping quickly. We build tooling, systems, frameworks, and systems to make sure everyone gets well configured, optimized developer environments.
Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Experience operating large‑scale clusters and container orchestration systems (e.g. Kubernetes or Slurm).
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Strong debugging across application, OS, and network layers.
Proficiency in Python or Rust (or similar), containers, and modern CI.
Experience with Kubernetes, controllers/operators, or performance profiling.
Familiarity with GPU/ML workflows or large‑scale data/eval pipelines.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're hiring an Executive Business Partner to support several technical leaders out of our San Francisco office. You will help our team stay focused and organized, managing personal logistics and any tasks that might fall through the cracks.
This is a non-traditional EA role, requiring creativity in adapting to different people’s work styles and the new challenges that emerge at a fast-moving startup. The role entails real autonomy in making decisions without tight supervision.
Manage calendars, schedule meetings, and coordinate travel for 3-4 technical leaders
Serve as the primary point of contact between your supported leaders and the rest of the company
Support recruiting coordination efforts
Track projects and commitments so nothing falls through the cracks
Minimum qualifications:
4+ years of executive or administrative support experience, ideally in AI labs or tech startups
Track record of supporting technical leaders such as engineers or researchers
Preferred qualifications:
Track record supporting a diverse team, adapting assistance and communication styles to individual needs
Experience managing a satellite office and coordinating across time zones
Experience adapting to a fast-changing role and meeting challenges proactively
Proven professionalism and discretion with sensitive or personal information
Having followed an executive to a new company
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $200,000 - $250,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We are hiring a developer productivity engineer to advance how we build software internally: safely, quickly, and with delight. The main focus are AI tools and coding agents. You’ll partner with platform, security, and product engineers to build state-of-the-art tooling for AI-assisted software development, and make our inner loop dramatically faster.
The scope of this role includes both setting up company-wide platforms and working with developers to accelerate their individual workflows.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Enable our researchers and engineers to leverage AI to improve coding productivity without compromising code quality
Standardize AI coding tools, such as Claude Code, Cursor, and Codex. You will help configure, harden, and maintain the best tools, integrating org-wide configurations with individual preferences.
Build secure, reproducible agent sandboxes for remote dev & CI testing.
Set up golden-path dev environments and guardrails for secrets/PII.
Help individual contributors develop their personalized AI-enabled workflow.
Track tool usage, reliability, and cost.
Minimum qualifications:
Bachelor’s degree or equivalent industry experience in computer science, engineering, or similar.
Experience developing productivity tools and best practices for large codebases.
Ability to communicate clearly and work with researchers to build and manage a variety of internal tools.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
Hands-on experience with container platforms (e.g. Docker/Kubernetes), modern CI (GitHub Actions/Buildkite), and package management tools (uv).
Practical experience with AI coding tools and model APIs (e.g. OSS via vLLM / SGLang / TGI).
Solid Linux/networking fundamentals; comfort with secrets management and safe egress.
Proficiency in systems programming languages (e.g. Rust) and scripting languages (e.g. Python).
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an infrastructure research engineer to design and build the core systems that enable efficient large-scale model training with a focus on numerics. You will focus on improving the numerical foundations of our distributed training stack, from precision formats and kernel optimizations to communication frameworks that make training trillion-parameter models stable, scalable, and fast.
This role is ideal for someone who thrives at the intersection of research and systems engineering: a builder who understands both the math of optimization and the realities of distributed compute.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Design and optimize distributed training infrastructure for large-scale LLMs, focusing on performance, stability, and reproducibility across multi-GPU and multi-node setups.
Implement and evaluate low-precision numerics (for example, BF16, MXFP8, NVFP4) to improve efficiency without sacrificing model quality.
Develop kernels and communication primitives that use hardware-level support for mixed and low-precision arithmetic.
Collaborate with research teams to co-design model architectures and training recipes that align with emerging numeric formats and stability constraints.
Prototype and benchmark scaling strategies such as data, tensor, and pipeline parallelism that integrate precision-adaptive computation and quantized communication.
Contribute to the design of our internal orchestration and monitoring systems to ensure that thousands of distributed experiments can run efficiently and reproducibly.
Publish and share learnings through internal documentation, open-source libraries, or technical reports that advance the field of scalable AI infrastructure.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, electrical engineering, statistics, machine learning, physics, robotics, or similar.
Understanding of deep learning frameworks (e.g., PyTorch, JAX) and their underlying system architectures.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Strong engineering skills, ability to contribute performant, maintainable code and debug in complex codebases in areas such as floating-point numerics, low-precision arithmetic, and distributed systems.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Familiarity with distributed frameworks such as PyTorch/XLA, DeepSpeed, Megatron-LM.
Experience implementing FP8, INT8, or block-floating point (MX) formats and understanding their numerical trade-offs.
Prior contributions to open-source deep learning infrastructure such as PyTorch, DeepSpeed, or XLA.
Publications, patents, or projects related to numerical optimization, communication-efficient training, or systems for large models.
Experience training and supporting large-scale AI models.
Track record of improving research productivity through infrastructure design or process improvements.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
The IT team builds secure infrastructure and efficient processes that enable our employees to move quickly. We operate an all-Mac environment and manage our endpoint fleet as a distributed platform, applying production-engineering practices to device management and security.
Our endpoint configurations, security policies, scripts, and software deployments are increasingly managed through version-controlled workflows with testing, review, staged rollouts, and rollback capabilities. This role will work closely with IT, Security, Identity, and Infrastructure to deliver a secure and reliable employee computing experience.
Endpoint Configuration as Code: Author, review, test, and progressively deploy macOS configuration profiles, security policies, queries, and remediation scripts. Build code review, staging, canary, validation, and rollback processes into endpoint changes.
MDM Platform Engineering: Operate our MDM platform as a production service, including configuration as code, observability, upgrades, reliability, incident response, and integrations with other IT and Security systems.
MDM Migration: Lead the evaluation, design, testing, and execution of our planned migration from Iru to Fleet. Establish functional requirements, identify configuration and security-control gaps, develop a phased migration plan, and move the fleet with minimal disruption to employees.
Santa and Rudolph: Own the architecture and operation of Santa and its Rudolph synchronization service. Manage binary-authorization policies, rule distribution, application approvals, telemetry, observability, infrastructure, and incident response.
Zero Trust and Device Trust: Partner closely with Security and Identity to make device trust a core component of our Zero Trust architecture. Integrate endpoint posture signals into authentication, authorization, and conditional-access decisions.
Continuous Posture Evaluation: Build systems that continuously evaluate device health and security posture, including MDM enrollment, OS version, patch status, disk encryption, endpoint protection, security-control status, and configuration compliance. Automatically identify and remediate drift or restrict access when a device no longer meets requirements.
Patch Management: Build and maintain automated macOS patching workflows that support rapid enforcement timelines while providing a thoughtful employee experience.
Zero-Touch Provisioning: Design and improve Apple Business Manager and Automated Device Enrollment workflows that turn a new Mac into a secure, fully configured, and productive machine with minimal manual intervention.
Software Distribution: Own application packaging, deployment, updating, and removal across the Mac fleet.
Fleet Telemetry and Compliance: Query live device state at scale and turn endpoint telemetry into actionable policies, dashboards, compliance reporting, and early warnings for configuration drift.
Automation: Build tools and AI-assisted workflows that reduce repetitive operational work and make endpoint management more reliable and scalable.
Endpoint Security: Partner with Security on macOS hardening, binary authorization, vulnerability management, compliance controls, detection and response, and device-based access policies.
Advanced Troubleshooting: Serve as the escalation point for complex macOS and endpoint-platform issues that cannot be resolved through standard IT support processes.
Technical Leadership: Help define the endpoint roadmap, evaluate technologies, make architecture decisions, and lead complex initiatives from conception through production.
8+ years of experience building and operating secure IT or endpoint systems in complex environments.
Experience managing a large fleet of macOS devices through a modern MDM platform.
Experience managing endpoint configuration through scripted deployments, Git-based workflows, or a full GitOps model.
Deep knowledge of macOS internals, enterprise deployment, security controls, and troubleshooting.
Experience designing and operating zero-touch Mac provisioning, patching, and software-distribution workflows.
Experience using device health and security signals to evaluate endpoint compliance.
Experience successfully delivering complex technical projects from conception through production.
Strong ability to solve ambiguous problems involving multiple teams and stakeholders.
Ability to communicate technical concepts clearly to technical and nontechnical audiences.
A product-engineering mindset toward IT systems, including testing, observability, reliability, and controlled change management.
A consistent practice of creating clear technical documentation, architecture diagrams, runbooks, and operational procedures.
Ability to work from either our New York or San Francisco office.
Fleet: Experience deploying, operating, or contributing to Fleet, including its MDM, osquery, GitOps, software-management, and vulnerability-management capabilities.
MDM Migration: Experience leading a production MDM migration, particularly in an environment using Apple Business Manager and Automated Device Enrollment.
Iru: Experience managing macOS devices with Iru, formerly Kandji.
Santa and Rudolph: Experience operating Santa at scale, including rule management, binary authorization, event telemetry, and a Rudolph synchronization service.
Zero Trust: Experience designing device-trust and continuous-posture-evaluation systems that integrate with identity providers, conditional access, or other Zero Trust controls.
MDM as a Service: Experience operating an MDM or device-management platform as a production service rather than only administering a SaaS console.
Progressive Delivery: Experience building automated endpoint rollout systems with staging, canary groups, rollback capabilities, and promotion decisions based on telemetry.
Open-Source Tooling: Experience deploying, operating, or contributing to open-source macOS endpoint-management or security tools.
Infrastructure as Code: Experience managing endpoint or cloud infrastructure through Terraform or another infrastructure-as-code framework.
Cloud Infrastructure: Experience operating AWS services such as Lambda, API Gateway, DynamoDB, containers, managed databases, and monitoring systems.
Endpoint Development: Proficiency in Swift or Go for building macOS endpoint tools, agents, or supporting services.
AI-Assisted Operations: Experience using LLMs to automate operational work or a strong interest in applying them to endpoint engineering.
Python and shell scripting.
macOS internals, including launchd, configuration profiles, Transparency, Consent, and Control (TCC), system extensions, Endpoint Security, FileVault, Secure Token, and bootstrap tokens.
Apple Business Manager, Automated Device Enrollment, and Apple’s MDM and Declarative Device Management frameworks.
Modern Apple MDM platforms, particularly Iru, Fleet, Jamf, or equivalent.
Santa binary authorization and Rudolph synchronization infrastructure.
Fleet-scale querying and osquery.
Git, pull-request workflows, GitOps, and CI/CD for endpoint configuration.
Terraform and infrastructure as code.
Public-cloud fundamentals, including serverless infrastructure, containers, managed databases, and monitoring.
Device lifecycle automation, including zero-touch enrollment, patching, software distribution, and secure deprovisioning.
Endpoint security, Zero Trust, device trust, continuous posture evaluation, compliance, and automated remediation.
Location: This role is based in San Francisco, California or New York, New York.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $180,000 - $360,000.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an infrastructure research engineer to design, optimize, and scale the systems that power large AI models. Your work will make inference faster, more cost-effective, more reliable, and more reproducible to enable our teams to focus on advancing model capabilities rather than managing bottlenecks.
Our focus is on performant and efficient model inference both to power real-world applications and to accelerate research. This role is responsible for the infrastructure that ensures every experiment, evaluation, and deployment runs smoothly at scale.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Work alongside researchers and engineers to bring cutting-edge AI models into production.
Collaborate with research teams to enable high-performance inference for novel architectures.
Design and implement new techniques, tools, and architectures that improve performance, latency, throughput, and efficiency.
Optimize our codebase and compute fleet (e.g., GPUs) to fully utilize hardware FLOPs, bandwidth, and memory.
Extend orchestration frameworks (e.g., Kubernetes, Ray, SLURM) for distributed inference, evaluation, and large-batch serving.
Establish standards for reliability, observability, and reproducibility across the inference stack.
Publish and share learnings through internal documentation, open-source libraries, or technical reports that advance the field of scalable AI infrastructure.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Understanding of deep learning frameworks (e.g., PyTorch, JAX) and their underlying system architectures.
Experience with inference serving systems optimized for throughput and latency (e.g., SGLang, vLLM).
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Strong engineering skills, ability to contribute performant, maintainable code and debug in complex codebases
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Experience training or supporting large-scale language models with hundreds of billions of parameters or more.
Understanding of distributed compute systems, GPU parallelism, and hardware-aware optimizations.
Contributions to open-source ML or systems infrastructure projects (e.g., SGLang, vLLM, PyTorch, Triton, DeepSpeed, XLA).
Track record of improving research productivity through infrastructure design or process improvements.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for a software engineer focused on making our products secure by default while supporting fast and ambitious product iteration. You’ll embed with product and research teams to bake security into design and development and to build tooling and automation that keep systems safe at scale.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Partner with product and research teams to embed security into the development lifecycle: threat modeling, design reviews, and secure defaults for new features.
Design and implement security controls across our product stack (authentication, authorization, session management, input validation, etc.).
Build and maintain security tooling and automation for engineers: secure frameworks and templates, CI/CD checks, dependency management, and vulnerability detection.
Collaborate with researchers to identify and mitigate AI-specific product risks, such as model abuse, prompt injection, data leakage, or misuse of capabilities.
Improve observability and detection for security-relevant events: access anomalies, abuse patterns, and suspicious behavior in production.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Strong generalist software engineering background and ability to review production code for security risks.
Hands-on experience securing web apps and APIs especially auth flows, access control, secrets management, input validation, and data protection.
Familiarity with common vulnerability classes and prevention frameworks; experience hardening prototypes into production.
Comfort with modern cloud infrastructure and understanding how application concerns intersect with infrastructure.
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Experience securing AI‑powered products or working with ML/LLM APIs and their unique threat models.
Background in human-computer interaction, especially where security or trust plays a central role in the user experience.
Strong skills in rapid prototyping and iteration, with a habit of turning ad-hoc fixes into reusable patterns and tools.
Open‑source security work, bug bounty write‑ups, or published tooling.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Thinking Machines builds multimodal-first. For us, there is no separate multimodal work. It’s at the core of everything we do, from the scientific goals we’re setting to the infrastructure we’re building. We’re looking for researchers to advance the frontier of audio capabilities. You’ll explore how audio models enable more natural and efficient communication/collaboration, preserving more information and capturing user intent.
This is a highly collaborative role. You’ll work closely across pre-training, post-training, and product with world-class researchers, infrastructure engineers, and designers. This is an opportunity to shape the fundamental capabilities of AI systems that millions of people will use.
This role blends fundamental research and practical engineering, as we do not distinguish between the two roles internally. You will be expected to write high-performance code and read technical reports. It’s an excellent fit for someone who enjoys both deep theoretical exploration and hands-on experimentation, and who wants to shape the foundations of how AI learns.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest in this research area. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Own research projects on audio training, low-latency inference and conversational responsiveness.
Design and train large-scale models that natively support audio input and output.
Investigate scaling behavior such as how data, model size, and compute affect capability and efficiency.
Build and maintain audio data pipelines, including preprocessing, filtering, segmentation, and alignment for training and evaluation.
Collaborate with data and infrastructure teams to scale audio training efficiently across distributed systems.
Publish and present research that moves the entire community forward. Share code, datasets, and insights that accelerate progress across industry and academia.
Minimum qualifications:
Ability to design, run, and analyze experiments thoughtfully, with demonstrated research judgment and empirical rigor.
Understanding of machine learning fundamentals, large-scale training, and distributed compute environments.
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
Experience with real-time inference, streaming architectures, or optimization for low latency.
Prior experience training or evaluating large-scale audio or multimodal models.
Publications, releases, or open-source projects related to speech, audio, voice, or similar areas.
Demonstrated experience in audio or speech modeling, including ASR, TTS, or self-supervised audio learning.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Thinking Machines Lab's mission is to empower humanity through advancing collaborative general intelligence. We're building a future where everyone has access to the knowledge and tools to make AI work for their unique needs and goals.
We are a small team of scientists, engineers, and builders who've created some of the most widely used AI products including ChatGPT, Character.ai, and PyTorch. As we scale our team, some of the hardest challenges we face are about empowering and aligning our people and helping our managers make crucial decisions under uncertainty.
The new role of HR Business Partner will combine two tasks: leadership coaching, and people systems design. The two require different skillsets, but a shared vision for managing talent.
You will coach managers at Thinking Machines Lab to be more effective leaders. You will provide strategic support on researcher and engineer performance, team dynamics, and personal growth. You will also build the people infrastructure that will scale this support as the company grows: performance and feedback systems, compensation structures, and career frameworks.
Coach managers by observing how they lead, identifying strengths and blind spots, and working on continuous improvement.
Advise leadership on organizational decisions: team structure, succession planning, and strategic people choices that shape how we work.
Design compensation structures that let us compete for the best machine learning talent in the world while staying aligned on values and principles.
Create career/leveling frameworks that work for a research lab where career advancement often doesn’t mean managing people, contributions such as mentorship and taste are harder to measure, and where senior researchers expect to grow and learn even after a decade in the role.
Build feedback and evaluation processes that help people improve, not just get measured.
Minimum qualifications:
5+ years experience working as a people leader or HR business partner in a high-growth technical environment.
Fluency with employment law and best practices in North America.
Proven track record of developing HR practices and supporting a diverse set of talent.
Preferred qualifications:
We encourage you to apply even if you don’t meet all preferred qualifications.
10+ years experience as a people leader or HR business partner.
Experience in setting up novel systems that scale in a fast-growing company.
Coaching experience that showcases a skill set in providing confidential support, motivating improvement, and conflict resolution.
Track record of supporting a high trust and low ego environment for high caliber talent with strong retention and growth.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $190,000 - $300,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're looking for a Site Reliability Engineer (SRE) to drive the reliability of Tinker end-to-end. You'll work alongside the engineers building the platform and research teams to make every layer of the system more robust and resilient.
Tinker is our fine-tuning API that empowers researchers and developers to customize frontier AI to their needs — opening access to capabilities that have previously been concentrated in a handful of labs. We manage the infrastructure while allowing Tinkerers full flexibility in training open weights models with their own data, algorithms, and for their own needs. Tinker is rapidly adding new customers, features, and novel use-cases. We’re hiring to grow the platform alongside the Tinker community.
Define and own end-to-end reliability, from CI/CD flows to production observability and incident response.
Develop appropriate Service Level Objectives for distributed training systems, balancing job completion reliability and scheduling latency with development velocity.
Design and implement monitoring and observability across the full training path.
Drive incident response for Tinker platform issues, ensuring rapid recovery, thorough incident reviews, and systematic improvements that prevent recurrence.
Harden multi-tenant isolation and resource scheduling so that LoRA-based workload co-scheduling maximizes utilization without compromising reliability or data separation
Collaborate with security teams to address production vulnerabilities
Bachelor's degree or equivalent experience in computer science, engineering, or similar.
Experience in distributed systems, cloud infrastructure, or site reliability engineering.
Proficiency writing software to solve reliability problems, including building tooling and automation.
Experience with production incident response, postmortems, and systematic reliability improvement.
Strong communication skills and track record of coordination across engineering and research teams.
Deep experience operating production cloud services at scale (e.g., public cloud platforms, internal cloud services)
Background in distributed training frameworks and how infrastructure failures surface in training behavior.
Track record building checkpoint and recovery systems for long-running distributed jobs.
Expertise in Kubernetes at scale: deploying, operating, debugging, and tuning clusters handling heterogeneous GPU workloads.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 – $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
The role of post-training researchers sits at the core of our roadmap. This is the critical bridge between raw model intelligence and a system that is actually useful, safe, and collaborative for humans.
This role blends fundamental research and practical engineering, as we do not distinguish between the two roles internally. You will be expected to write high-performance code and read technical reports. It’s an excellent fit for someone who enjoys both deep theoretical exploration and hands-on experimentation, and who wants to shape the foundations of how AI learns.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest in this research area. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Develop and tune the recipe: iterate on post-training recipes, consisting of a collection of datasets, training stages, and hyperparameters. Measure how recipe choices affect various metrics.
Iterate on evals: post-training involves a never-ending loop of defining a set of evaluations, optimizing them, and then realizing your existing evals don’t capture what matters. You’ll be responsible for both making numbers go up, and making sure the numbers are meaningful.
Debug and understand: while tuning the details of a training configuration, we often observe results that don’t quite make sense. You’ll be responsible for both getting things to work, and developing a deeper understanding, which we can bring to the next problem.
Scale and explore: post-training will involve a combination of scaling the existing methodologies and developing new ones. We’ll want to both measure how performance metrics scale with dataset size, and explore using a completely different kind of training dataset.
Publish and present research that moves the entire community forward. Share code, datasets, and insights that accelerate progress across industry and academia.
Minimum qualifications:
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
Prior experience with RLHF, RLAIF, preference modeling, or reward learning for large models.
Experience managing or analyzing human data collection campaigns or large-scale annotation workflows.
Research or engineering contributions in alignment, data-centric AI, or human-AI collaboration.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an infrastructure research engineer to design and build the core systems that enable scalable, efficient training of large models for deployment and research. Your goal is to make experimentation and training at Thinking Machines fast and reliable to ensure our research teams can focus on science, not system bottlenecks.
This role is ideal for someone who blends deep systems and performance expertise with a curiosity for machine learning at scale. You’ll take ownership of the training stack end to end, ensuring every GPU cycle drives scientific progress.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Design, implement, and optimize distributed training systems that scale across thousands of GPUs and nodes for large-scale training workloads.
Develop high-performance optimizations to maximize throughput and efficiency.
Develop reusable frameworks and libraries to improve training reproducibility, reliability, and scalability for new model architectures.
Establish standards for reliability, maintainability, and security, ensuring systems are robust under rapid iteration.
Collaborate with researchers and engineers to build scalable infrastructure.
Publish and share learnings through internal documentation, open-source libraries, or technical reports that advance the field of scalable AI infrastructure.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, electrical engineering, statistics, machine learning, physics, robotics, or similar.
Strong engineering skills, ability to contribute performant, maintainable code and debug in complex codebases
Understanding of deep learning frameworks (e.g., PyTorch, JAX) and their underlying system architectures.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Past experience working on distributed training for the world’s largest models to make them stable, reliable, and performant.
Track record of improving research productivity through infrastructure design or process improvements.
Contributions to open-source ML infrastructure such as PyTorch, XLA, Megatron-LM, or DeepSpeed.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an infrastructure research engineer to design, optimize, and maintain the compute foundations that power large-scale language model training. You will develop high-performance ML kernels (e.g., CUDA, CuTe, Triton), enable efficient low-precision arithmetic, and improve the distributed compute stack that makes training large models possible.
This role is perfect for an engineer who enjoys working close to the metal and across the research boundary. You’ll collaborate with researchers and systems architects to bridge algorithmic design with hardware efficiency. You’ll prototype new kernel implementations, profile performance across hardware generations, and help define the numerical and parallelism strategies that determine how we scale next-generation AI systems.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Design and implement custom ML kernels (e.g., CUDA, CuTe, Triton) for core LLM operations such as attention, matrix multiplication, gating, and normalization, optimized for modern GPU and accelerator architectures.
Design and think through compute primitives to reduce memory bandwidth bottlenecks and improve kernel compute efficiency.
Collaborate with research teams to align kernel-level optimizations with model architecture and algorithmic goals.
Develop and maintain a library of reusable kernels and performance benchmarks that serve as the foundation for internal model training.
Contribute to infrastructure stability and scalability, ensuring reproducibility, consistency across precision formats, and high utilization of compute resources.
Document and share insights through internal talks, technical papers, or open-source contributions to strengthen the broader ML systems community.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, electrical engineering, statistics, machine learning, physics, robotics, or similar.
Strong engineering skills, ability to contribute performant, maintainable code and debug in complex codebases
Understanding of deep learning frameworks (e.g., PyTorch, JAX) and their underlying system architectures.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Proficiency in CUDA, CuTe, Triton, or other GPU programming frameworks.
Demonstrated ability to analyze, profile, and optimize compute-intensive workloads.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Experience training or supporting large-scale language models with tens of billions of parameters or more.
Track record of improving research productivity through infrastructure design or process improvements.
Experience developing or tuning kernels for deep learning frameworks such as PyTorch, JAX, or custom accelerators.
Familiarity with tensor parallelism, pipeline parallelism, or distributed data processing frameworks.
Experience implementing low-precision formats (FP8, INT8, block floating point) or contributing to related compiler stacks (e.g., XLA, TVM).
Contributions to open-source GPU, ML systems, or compiler optimization projects.
Prior research or engineering experience in numerical optimization, communication-efficient training, or scalable AI infrastructure.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an engineer to design, build, and operate the GPU supercomputing environment that powers large‑scale training and inference. You will deliver high‑performant, reliable, and cost‑efficient compute so our users and researchers can move fast at scale.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Operate and automate large GPU clusters including provisioning, imaging, and capacity planning.
Write software that abstracts cluster management and presents a unified interface for training and inference.
Extend scheduling/orchestration (Kubernetes, Slurm, or similar) for topology‑aware placement, preemption, quotas, and fair‑share multi‑tenancy.
Monitor and improve operational metrics of speed, reliability, and error recovery.
Build reliable storage and artifact paths for datasets, checkpoints, and logs with clear retention and lineage.
Partner with researchers to unblock scale runs and advise on parallelism and performance trade‑offs.
Minimum qualifications:
Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Experience operating large‑scale clusters and container orchestration systems (e.g. Kubernetes or Slurm).
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Strong systems background: Linux, networking, and infrastructure‑as‑code.
Familiarity with CUDA/NCCL and performance profiling for distributed training/inference.
Prior work supporting large‑scale model training or inference environments.
Understanding of deep learning frameworks (e.g., PyTorch, TensorFlow, JAX) and their underlying system architectures.
Track record of working in fast-paced environments balancing care with urgency.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
The role of pre-training researchers sits at the core of our roadmap. This work blends research with large-scale data engineering to help assemble the pre-training datasets and data systems that underpin the next generation of AI models. You’ll design and implement methods for sourcing, curating, and analyzing pre-training data for quality and performance.
You’ll work with automated pipelines and human-in-the-loop processes, contributing both scientific insight and production-grade code. It’s ideal for someone who enjoys working at the intersection of data, machine learning, and systems, and who’s excited by the challenge of shaping frontier AI.
This role blends fundamental research and practical engineering, as we do not distinguish between the two roles internally. You will be expected to write high-performance code and read technical reports. It’s an excellent fit for someone who enjoys both deep theoretical exploration and hands-on experimentation, and who wants to shape the foundations of how AI learns.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest in this research area. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Design and implement techniques for curating, sourcing, and filtering large-scale text, code, and multimodal data.
Develop data quality metrics and analysis to measure coverage, diversity, and representativeness across sources.
Collaborate with research and infrastructure teams to scale data processing systems efficiently and reproducibly.
Investigate and mitigate data risks, including privacy, safety, and licensing concerns, to ensure responsible and ethical data use.
Continuously evaluate dataset improvements by analyzing their downstream effects on model learning and behavior.
Publish and present research that moves the entire community forward. Share code, datasets, and insights that accelerate progress across industry and academia.
Minimum qualifications:
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
Experience with curation, preprocessing, and analysis of large-scale text, code, or multimodal datasets.
Prior experience in data engineering, dataset construction, or large-scale web data processing for machine learning models.
Experience evaluating or improving training data quality and knowledge of data ethics, safety, and licensing frameworks relevant to AI dataset creation.
Contributions to open datasets, research publications, or data tooling.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're hiring a Recruiting Sourcer to build the pipeline of exceptional research and engineering talent that will define the next generation of AI. You'll partner closely with recruiters, hiring managers, and our technical leaders to identify, engage, and build relationships with candidates across research, engineering, and infrastructure.
This role is foundational to our growth. You'll need to think creatively about where to find rare technical talent, craft outreach that resonates with researchers and engineers who aren't actively looking, and help us build a sourcing function from the ground up at a fast-moving startup.
Build and manage a pipeline of candidates across research, engineering, and infrastructure roles, using creative and diverse sourcing channels
Partner with recruiters and hiring managers to deeply understand role requirements and translate them into effective sourcing strategies
Write and send personalized outreach that engages passive candidates, including senior researchers and engineers
Track pipeline metrics and sourcing effectiveness, and iterate on strategy based on what's working
Represent Thinking Machines' mission and culture authentically to prospective candidates throughout the sourcing process
Help build and refine our sourcing tools, processes, and infrastructure as the team scales
Minimum qualifications:
3+ years of sourcing or recruiting experience, ideally supporting technical roles in AI labs or tech startups
Track record of successfully identifying and engaging passive technical candidates, including researchers and engineers
Excellent written communication skills, with the ability to craft compelling, personalized outreach
Preferred qualifications:
Familiarity with the AI/ML research landscape, including key labs, conferences, and communities
Experience using sourcing tools and platforms (e.g., LinkedIn Recruiter, GitHub, academic search tools)
Experience building sourcing processes and infrastructure at an early-stage or fast-scaling company
Comfort operating with significant autonomy and adapting quickly as priorities shift
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $250,000 - $300,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an infrastructure engineer to own and evolve the security infrastructure that underpins our foundation models. In this role, you’ll work across compute, storage, networking, and data platforms, making sure our systems are secure, reliable, and built to scale. You’ll shape controls, architecture, and tooling so that security is part of how the platform works by default. You’ll partner closely with research and product teams, enabling them to move quickly while keeping our models, data, and environments protected.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Architect security patterns for platforms and services, including network segmentation, service-to-service authentication, RBAC, and policy enforcement in Kubernetes and cloud environments.
Manage identity, access, and secrets for humans and services: workload and cross-cloud identity, least-privilege IAM, and secrets management.
Build secure platforms for data ingestion, processing, and curation: classification, encryption, access controls, and safe sharing patterns across teams.
Write threat models and review designs with researchers and engineers to help them ship features and experiments in a safe, scalable way.
Automate security checks and build guardrails: policy-as-code, secure infrastructure baselines, validation in CI/CD, and tools that make the secure path the easiest one.
Minimum qualifications:
Bachelor’s degree or equivalent experience in engineering, or similar.
Strong background with containers and orchestration (e.g., Kubernetes) and how to secure them (namespaces, network policies, pod security, admission controls, etc.)
Practical experience with Infrastructure as Code (Terraform or similar), including secure patterns for provisioning networks, IAM, and shared services.
Solid understanding of cloud networking and security: VPCs, load balancers, service discovery, mTLS, firewalls, and zero-trust-style architectures.
Proficiency with a systems language such as Rust and scripting in Python for building platform components and internal tools.
Evidence of owning complex, production-critical systems, including debugging issues that span infra, security, and application layers.
Preferred qualifications — we encourage you to apply if you meet some even if you don't meet all of these:
Experience with ML infrastructure, GPU clusters, or large-scale training environments (schedulers, job queues, shared storage, multi-tenant clusters).
Background in AI labs, HPC environments, or ML-heavy organizations where both security and performance are first-class concerns.
Experience profiling and tuning high-throughput systems, and an ability to reason about the cost of additional security layers.
Talks, blogs, or publications on infrastructure security, distributed systems, or performance engineering.
Open-source contributions to security, orchestration, observability, or infrastructure tooling.
Familiarity with securing specialized hardware (GPUs, TPUs) and their integrations into training and inference pipelines.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $200,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
As Product Manager for Deployment, you will own how Thinking Machines' models and fine-tuned checkpoints go from training into production use. You will shape the path from a trained model to a served, reliable, cost-effective endpoint — covering inference infrastructure, serving APIs, latency and throughput tradeoffs, scaling behavior, observability, and the workflows researchers and external users rely on to deploy their work with confidence.
This is not a mature MLOps role at an established platform. Deployment at Thinking Machines is still being defined: what \"production-ready\" means for a fine-tuned model, which serving paths we support, how much control users get over performance and cost tradeoffs, and how we scale reliably as usage grows. You will work from infrastructure capability through to a deployment experience that is fast, predictable, and trustworthy.
The strongest candidate has shipped and operated production ML or infrastructure systems before, ideally as an engineer before becoming a product leader, and can reason from strategy down to autoscaling behavior, latency budgets, rollout safety, and the on-call realities of running models in production.
What You'll Do
Own deployment strategy, roadmap, and success metrics for taking models and Tinker-trained checkpoints into production, in close partnership with infrastructure, research, engineering, and GTM
Define priority deployment paths and workflows across model serving, autoscaling, versioning, rollback, monitoring, and incident response
Work at engineering depth on serving architecture, latency and cost tradeoffs, reliability targets, capacity planning, and API/SDK surfaces for deployment
Build direct feedback loops with users deploying models in production, and turn scattered signals into a clear view of what's broken, what's missing, and what to prioritize next
Drive ambiguous workstreams end to end: technical scoping, dependency resolution, launch readiness, on-call/escalation design, and post-incident learning
Connect deployment decisions to the model and infrastructure roadmap, making visible the tradeoffs between flexibility, reliability, and operational cost
Shape SLAs, pricing/packaging inputs for hosted inference, and the operating model for a deployment platform expected to scale quickly
Do whatever work makes deployment succeed — reviewing a serving config, joining an incident retro, inspecting latency data, or writing the rollout plan for a new model
Skills and Qualifications
Experience owning a production ML serving, infrastructure, or deployment product, with direct involvement in reliability, scaling, or performance decisions
Track record working at engineering depth with production systems — comfortable discussing latency, throughput, autoscaling, rollback, or incident response in specifics
Experience taking a technical product from early usage through to reliable, scaled production use
Preferred qualifications:
Background as an engineer or technical founder before moving into product leadership
Experience with ML inference infrastructure specifically (model serving frameworks, GPU scheduling, batching, quantization tradeoffs, or similar)
Experience operating in a startup, lab, or new product area where the deployment model and roadmap weren't handed to you
Comfortable moving between a strategic narrative and a specific technical detail (an autoscaling policy, an SLA definition, a rollout gate) without losing judgment
Experience building trust with technical users through evidence, responsiveness, and follow-through rather than process ownership
Logistics
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $450,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're hiring an Engineering Manager to lead a team of senior and staff-level engineers across ML infrastructure and product. You will help the team build and scale systems that are reliable, performant, and easy to operate.
This role combines collaboration with hand-on work. You’ll partner with tech leads to set the technical direction for your team and own its execution. You should also be ready to go deep on system design and contribute directly when needed.
Lead and grow a team of senior and staff-level engineers, setting clear expectations and maintaining a high bar for execution.
Own architecture, system design, and long-term technical direction for your team's systems, with emphasis on reliability and performance.
Contribute directly to design reviews, prototyping, and debugging critical issues.
Partner with researchers and product teams to define roadmaps and prioritize work.
Hire and close senior engineering talent. Mentor engineers into technical leaders.
Bachelor’s degree or equivalent industry experience in computer science, engineering, or similar.
8+ years of experience building and scaling production systems, including system design and distributed systems.
3+ years of engineering management experience in high-growth environments.
Experience managing teams of senior or staff-level engineers.
Background in infrastructure, systems engineering, or developer productivity.
Familiarity with AI/ML systems, data infrastructure, or high-performance computing.
Track record of building or contributing to widely used systems, platforms, or tools.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $400,000 - $500,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
The IT team builds secure infrastructure and efficient processes that enable our employees to move quickly. We operate an all-Mac environment and use tools such as Okta, Google Workspace, and Kandji to support a rapidly growing workforce across multiple offices.
This role will serve as the primary IT partner for our New York employees and office. You will work closely with the IT and Security team in San Francisco to maintain consistent systems, security standards, and employee experiences across locations.
New York IT Support: Serve as the primary onsite IT resource for employees in our New York office, providing responsive support and resolving hardware, software, access, network, and workplace technology issues.
Employee Lifecycle Management: Own onboarding and offboarding for New York employees, including hardware procurement, device preparation, shipping logistics, account provisioning, and access changes.
Identity and Access Management: Administer Okta, including Single Sign-On, application integrations, provisioning, and access management for new and existing services.
Infrastructure as Code: Help manage Okta and other SaaS platforms through Terraform and Git-based workflows. Contribute to the team’s broader effort to manage corporate technology systems using repeatable, reviewable, and version-controlled processes.
Office Infrastructure and Security: Own the day-to-day operation of the New York office’s technology infrastructure, including ISPs, networking, firewalls, conference rooms, and physical security systems such as door access and cameras.
A/V and Events: Set up, maintain, and support conference room technology, hybrid meetings, and company-wide events such as All Hands meetings.
Device Fleet Management: Maintain and secure the company’s fleet of macOS devices using MDM, ensuring devices remain compliant, updated, and reliable.
Cross-Office Collaboration: Partner closely with the San Francisco IT and Security team on company-wide initiatives, incident response, standards, documentation, and infrastructure improvements.
Vendors and Projects: Coordinate with vendors, contractors, and building management on office technology installations, upgrades, and repairs.
5+ years of experience in IT engineering, systems administration, corporate IT, or a similar role.
Experience supporting employees and technology infrastructure in an onsite office environment.
Strong experience administering identity, SaaS, endpoint-management, and collaboration platforms.
Working knowledge of Git and version-controlled operational workflows.
Ability to independently own IT operations for the New York office while collaborating effectively with a distributed team.
Ability to work onsite Monday through Friday in our New York office.
Ability to travel to San Francisco for approximately one week each month.
SaaS Administration: Deep experience administering Google Workspace, Slack Enterprise, GitHub, and similar business platforms.
Identity and Security: Strong knowledge of Okta, SSO, application provisioning, access-management practices, and Zero Trust security principles.
Terraform and GitOps: Experience managing Okta or other SaaS platforms with Terraform. Familiarity with pull requests, code reviews, CI/CD, state management, and GitOps practices is strongly preferred.
Apple Device Management: Experience managing macOS devices through a modern MDM platform. Kandji experience is a plus.
Networking: Experience supporting office networks, wireless infrastructure, ISPs, firewalls, and related troubleshooting.
Physical Security: Familiarity with access-control and camera platforms such as Brivo.
Automation: Ability to automate operational processes using Python, Bash, APIs, or similar tools.
Project Leadership: Experience managing vendors and leading office technology projects, particularly those involving external contractors, office openings, or infrastructure upgrades.
Documentation: A track record of creating clear technical documentation, runbooks, and repeatable operational processes.
Location: This role is based in New York, New York.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $180,000 - $360,000.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Thinking Machines builds multimodal-first. We’re looking for new team members to advance the science of visual perception and multimodal learning. We think about how vision and language interact at scale. We design architectures that fuse pixels and text, build datasets and evaluation methods that test real-world comprehension, and develop representations that let models ground abstract concepts in the physical world. Our goal is to create multimodal systems that support seamless integration into real-world environments.
You’ll work at the intersection of visual understanding, multimodal reasoning, and large-scale model training. You’ll help develop the architectures, data, and evaluation tools that teach AI to see, understand, and collaborate. The best candidate is curious about multimodal interfaces, has experience running large scale experiments and is comfortable contributing to complex engineering systems. While we are looking for a person with expertise in multimodality, Thinking Machines Lab operates in a unified fashion and expects new hires to work across modalities as one team.
This role blends fundamental research and practical engineering, as we do not distinguish between the two roles internally. You will be expected to write high-performance code and read technical reports. It’s an excellent fit for someone who enjoys both deep theoretical exploration and hands-on experimentation, and who wants to shape the foundations of how AI learns.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest in this research area. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Own research projects on training and performance analysis of multimodal AI models.
Curate and build large-scale datasets and evaluation benchmarks to advance vision capabilities.
Work with our data infrastructure engineers, pretraining researchers and engineers, and product team to create frontier multimodal models and the products that leverage them.
Publish and present research that moves the entire community forward. Share code, datasets, and insights that accelerate progress across industry and academia.
Minimum qualifications:
Ability to design, run, and analyze experiments thoughtfully, with demonstrated research judgment and empirical rigor.
Understanding of machine learning fundamentals, large-scale training, and distributed compute environments.
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
Research or engineering contributions in visual reasoning, spatial understanding, or multimodal architecture design.
Experience developing evaluation frameworks for multimodal tasks.
Publications or open-source contributions in vision-language modeling, video understanding, or multimodal AI.
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're looking for a GRC Lead who personally drives our certifications (SOC 2, ISO 27001, FedRAMP and others as we grow) from scoping through audit close, and runs our compliance processes day to day. You'll collect the evidence, write the control documentation, and sit across from the auditor yourself.
You'll work closely with security, legal, safety, and engineering to answer compliance and risk questions directly, using your own technical understanding of how our systems work. Day to day, you'll be managing audits, controls, and risk assessments. Alongside that, you'll be building the roadmap for what this function needs to look like in a year.
Own our certification roadmap end to end: scope each certification, build the control set, collect and organize evidence, and represent TML directly to auditors through to close.
Manage recurring compliance processes on a set cadence: control testing, audit prep and response, risk register maintenance, and policy attestations.
Answer compliance and risk questions from engineering, security, and product teams directly, by building enough technical fluency across our infrastructure, model deployment, and data handling to do so without escalating every question.
Track regulatory and framework requirements relevant to an AI company (GDPR, EU AI Act, and similar) and translate them into specific, actionable controls.
Identify gaps in current compliance coverage as the company adds new products, infrastructure, or jurisdictions, and propose what needs to change before it becomes a blocker.
Build and maintain the tooling and documentation that make the next audit cycle faster than the last one.
Plan a multi-quarter roadmap for the GRC function itself, while continuing to personally run the certifications and audits already on the books.
Minimum qualifications:
7+ years related experience across technology and cybersecurity Governance, Risk, and Compliance (GRC), with demonstrated breadth across all three disciplines.
Experience leading a SOC 2, ISO 27001, FedRAMP or comparable certification from scoping through audit close.
Hands-on experience collecting audit evidence and writing control documentation.
Experience managing a recurring compliance process, such as control testing, risk register maintenance, or policy attestations.
Experience learning new technical domains quickly and translating them for non-technical stakeholders.
Preferred qualifications:
We encourage you to apply even if you don't meet all preferred qualifications.
Background as a software engineer or in a technical engineering role, now applied to GRC, evidenced by scripts, tools, or automations you've personally built for evidence collection, control testing, or audit workflows.
Experience translating complex compliance requirements into scalable automation using AI agents and custom built tooling.
Experience growing a GRC function's capability (new certifications, tooling, or processes) as a company scaled.
You want to run the certification yourself, end to end.
You're the one in the room with the auditor, walking through evidence.
You can hold this week's deadlines and next year's roadmap at the same time.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $225,000 - $350,000.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
As a safety researcher, you'll work toward ensuring our models are safe and trustworthy. The role sits at the intersection of research and hands-on technical work. A central question is how models come to handle harmful or dual-use requests: what they learn from data, how training shapes where they refuse and where they engage, and what makes those boundaries reliable. You'll explore the science behind these behaviors and design experiments that inform how our models are trained and evaluated.
We are hiring across the entire development stack — from pre-training data curation to safety-focused fine-tuning, evaluations, and red-teaming. During project selection we’ll take into account your interests and experience alongside organizational needs. This flexible approach allows us to match talented safety researchers with the teams where they'll have the greatest impact and growth potential.
Here are example areas you may contribute to depending on your area of expertise and interest:
Build data filtering pipelines and quality classifiers to shape what models learn from pre-training corpora, and study how those early interventions affect downstream safety behavior.
Apply post-training techniques, including RL from human and AI feedback and policy-based reasoning approaches, to shape how models handle harmful, sensitive, and dual-use requests.
Design, build, and maintain safety evaluations, with particular focus on measuring model behavior on long-horizon and agentic tasks.
Generate and curate synthetic data to train and evaluate models on refusal boundaries and safety-relevant behaviors.
Red-team our models and products to surface failure modes, jailbreaks, and emergent risks before deployment, and design mitigations for what you find.
Required qualifications:
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Background in AI safety research, with hands-on experience in at least one area of safety, such as: RLHF/RLAIF, alignment and preference modeling, deliberative alignment, safety evaluations, or red-teaming.
Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Experience building evaluations for long-horizon, multi-step, or agentic tasks.
Experience generating synthetic data at scale for training or evaluation.
Experience with modern red-teaming/jailbreaking techniques.
Research contributions in AI safety — publications, open-source evaluations, or public red-teaming work.
Familiarity with the AI safety literature and current open problems (e.g., scalable oversight, reward hacking, jailbreak robustness).
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nAs a safety researcher, you'll work toward ensuring our models are safe and trustworthy. The role sits at the intersection of research and hands-on technical work. A central question is how models come to handle harmful or dual-use requests: what they learn from data, how training shapes where they refuse and where they engage, and what makes those boundaries reliable. You'll explore the science behind these behaviors and design experiments that inform how our models are trained and evaluated.\n\n\n\n\nWHAT YOU’LL DO\n\nWe are hiring across the entire development stack — from pre-training data curation to safety-focused fine-tuning, evaluations, and red-teaming. During project selection we’ll take into account your interests and experience alongside organizational needs. This flexible approach allows us to match talented safety researchers with the teams where they'll have the greatest impact and growth potential.\n\nHere are example areas you may contribute to depending on your area of expertise and interest:\n\n - Build data filtering pipelines and quality classifiers to shape what models learn from pre-training corpora, and study how those early interventions affect downstream safety behavior.\n\n - Apply post-training techniques, including RL from human and AI feedback and policy-based reasoning approaches, to shape how models handle harmful, sensitive, and dual-use requests.\n\n - Design, build, and maintain safety evaluations, with particular focus on measuring model behavior on long-horizon and agentic tasks.\n\n - Generate and curate synthetic data to train and evaluate models on refusal boundaries and safety-relevant behaviors.\n\n - Red-team our models and products to surface failure modes, jailbreaks, and emergent risks before deployment, and design mitigations for what you find.\n\n\n\n\nSKILLS AND QUALIFICATIONS\n\nRequired qualifications:\n\n - Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.\n\n - Background in AI safety research, with hands-on experience in at least one area of safety, such as: RLHF/RLAIF, alignment and preference modeling, deliberative alignment, safety evaluations, or red-teaming.\n\n - Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.\n\n - Clarity in communication, an ability to explain complex technical concepts in writing.\n\nPreferred qualifications — we encourage you to apply if you meet some but not all of these:\n\n - Experience building evaluations for long-horizon, multi-step, or agentic tasks.\n\n - Experience generating synthetic data at scale for training or evaluation.\n\n - Experience with modern red-teaming/jailbreaking techniques.\n\n - Research contributions in AI safety — publications, open-source evaluations, or public red-teaming work.\n\n - Familiarity with the AI safety literature and current open problems (e.g., scalable oversight, reward hacking, jailbreak robustness).\n\n - PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.\n \n \n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."},{"id":"0871bc29-7ca0-4906-a674-7b32527617ab","title":"Research, RL Scaling","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-08-21T19:15:28.384+00:00","isListed":true,"isRemote":true,"workplaceType":"Hybrid","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/0871bc29-7ca0-4906-a674-7b32527617ab","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/0871bc29-7ca0-4906-a674-7b32527617ab/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Our team scales reinforcement learning for frontier models. Progress in RL is increasingly set by how well it scales: more rollouts, larger models, and training loops that keep large fleets of accelerators doing useful work. We are particularly interested in people working on high-training-compute, long-horizon RL. We believe the biggest gains come from designing the training recipe and the infrastructure together rather than separately, and we are hiring a researcher who wants to own that boundary.
A center of gravity for this role is asynchronous RL. Decoupling generation from training changes both the systems design and the learning problem, and doing it well requires a deep understanding of async RL algorithms, design choices, and trade-offs on both the ML and the systems sides. We expect much of the headroom in RL scaling to come from here.
Because generation dominates the cost of RL at scale, good knowledge of inference systems, low-precision numerics, and quantization is recommended: you should be able to reason quantitatively about rollout throughput and cost (batching, KV cache, MoE serving, speculative decoding) and about how inference constraints shape training design.
This is a research role with full-stack ownership, from the algorithms to the parallelism plan to the health of the run.
Co-design the RL recipe and the systems that run it: make recipe-level choices jointly with systems-level ones and validate them at frontier scale.
Advance asynchronous RL algorithms.
Improve the efficiency of rollout generation and its integration with training, treating inference as a first-class part of the RL loop.
Run frontier-scale RL end to end: bring up new models and training setups, keep large runs stable and healthy.
Jointly optimize the compute and training efficiency of RL: accelerator utilization, memory, communication, and low-precision numerics.
Do careful empirical science: ablations and scaling studies backed by instrumentation you can trust, written up clearly.
Minimum qualifications:
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Strong research judgment: clean ablations, honest baselines, and clear technical writing.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Strong grounding in RL for large language models, such as modern policy optimization methods and their behavior at scale.
Deep understanding of asynchronous RL: the algorithms, design choices, and trade-offs, on both the ML and the systems sides.
Experience training large models across many accelerators, with comfort inside the distributed stack (parallelism strategies, memory, communication).
Good working knowledge of inference systems: able to reason quantitatively about rollout generation throughput and cost.
Experience building or operating decoupled generation/training RL systems at scale.
Experience with RL on verifiable and agentic tasks, including multi-turn environments.
Experience with RL training stability techniques for large runs.
Familiarity with low-precision training and inference: numerics, quantization, and their implications for RL.
Hands-on work with LLM serving stacks (e.g., SGLang, vLLM, TokenSpeed, or custom engines).
Experience with scaling studies for large models.
Contributions to open-source training or inference frameworks.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nOur team scales reinforcement learning for frontier models. Progress in RL is increasingly set by how well it scales: more rollouts, larger models, and training loops that keep large fleets of accelerators doing useful work. We are particularly interested in people working on high-training-compute, long-horizon RL. We believe the biggest gains come from designing the training recipe and the infrastructure together rather than separately, and we are hiring a researcher who wants to own that boundary.\n\nA center of gravity for this role is asynchronous RL. Decoupling generation from training changes both the systems design and the learning problem, and doing it well requires a deep understanding of async RL algorithms, design choices, and trade-offs on both the ML and the systems sides. We expect much of the headroom in RL scaling to come from here.\n\nBecause generation dominates the cost of RL at scale, good knowledge of inference systems, low-precision numerics, and quantization is recommended: you should be able to reason quantitatively about rollout throughput and cost (batching, KV cache, MoE serving, speculative decoding) and about how inference constraints shape training design.\n\nThis is a research role with full-stack ownership, from the algorithms to the parallelism plan to the health of the run.\n\n\n\n\nWHAT YOU’LL DO\n\n - Co-design the RL recipe and the systems that run it: make recipe-level choices jointly with systems-level ones and validate them at frontier scale.\n\n - Advance asynchronous RL algorithms.\n\n - Improve the efficiency of rollout generation and its integration with training, treating inference as a first-class part of the RL loop.\n\n - Run frontier-scale RL end to end: bring up new models and training setups, keep large runs stable and healthy.\n\n - Jointly optimize the compute and training efficiency of RL: accelerator utilization, memory, communication, and low-precision numerics.\n\n - Do careful empirical science: ablations and scaling studies backed by instrumentation you can trust, written up clearly.\n\n\n\n\nSKILLS AND QUALIFICATIONS\n\nMinimum qualifications:\n\n - Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.\n\n - Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.\n\n - Clarity in communication, an ability to explain complex technical concepts in writing.\n\n - Strong research judgment: clean ablations, honest baselines, and clear technical writing.\n\n \n\nPreferred qualifications — we encourage you to apply if you meet some but not all of these:\n\n - PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.\n\n - Strong grounding in RL for large language models, such as modern policy optimization methods and their behavior at scale.\n\n - Deep understanding of asynchronous RL: the algorithms, design choices, and trade-offs, on both the ML and the systems sides.\n\n - Experience training large models across many accelerators, with comfort inside the distributed stack (parallelism strategies, memory, communication).\n\n - Good working knowledge of inference systems: able to reason quantitatively about rollout generation throughput and cost.\n\n - Experience building or operating decoupled generation/training RL systems at scale.\n\n - Experience with RL on verifiable and agentic tasks, including multi-turn environments.\n\n - Experience with RL training stability techniques for large runs.\n\n - Familiarity with low-precision training and inference: numerics, quantization, and their implications for RL.\n\n - Hands-on work with LLM serving stacks (e.g., SGLang, vLLM, TokenSpeed, or custom engines).\n\n - Experience with scaling studies for large models.\n\n - Contributions to open-source training or inference frameworks.\n\n\n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."},{"id":"6afeaa7a-8674-4e32-9364-fb14cc28e79a","title":"Software Engineer, Sandboxing","department":"Core Engineering","team":"Core Engineering","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-08-26T00:01:05.844+00:00","isListed":true,"isRemote":true,"workplaceType":"Hybrid","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/6afeaa7a-8674-4e32-9364-fb14cc28e79a","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/6afeaa7a-8674-4e32-9364-fb14cc28e79a/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Our models and agents increasingly need to run code, use tools, and take actions in the world — safely, reliably, and at scale. The Core Services team builds the sandboxing infrastructure that makes this possible: the isolated execution environments where models write and run code, browse, and interact with tools, both for our researchers during training and for external users building on Tinker.
We're hiring a software engineer to help design, build, and operate this sandboxing platform. You'll work on the systems that isolate and constrain untrusted, model-generated code, and that scale to support thousands of concurrent executions across the company. This is foundational infrastructure: every research experiment and every product surface that lets a model take action depends on it being fast, secure, and dependable.
Design, build, and operate sandboxed execution environments for running untrusted, model-generated code and tool calls at scale.
Improve isolation boundaries using technologies such as containers, microVMs, or gVisor-style kernels, balancing security against startup latency and throughput.
Build the scheduling, resource-management, and lifecycle systems that provision, reuse, and tear down sandboxes efficiently under heavy concurrent load.
Partner with researchers and Tinker's product team to expose sandboxing primitives that are simple to use and hard to misuse.
Instrument sandboxes for observability and abuse detection, and respond to novel escape or exploitation attempts as they're discovered.
Own reliability and performance of the sandboxing platform end-to-end, from API design down to the underlying virtualization layer.
Minimum qualifications:
Bachelor's degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Experience building or operating isolation or virtualization technology, such as containers, microVMs (e.g. Firecracker, Cloud Hypervisor), or sandboxed runtimes (e.g. gVisor, Kata Containers).
Solid grounding in Linux internals relevant to isolation: namespaces, cgroups, seccomp, capabilities, and networking.
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
Preferred qualifications:
Experience securing systems that execute untrusted or adversarial code, including threat modeling and hardening against sandbox escapes.
Familiarity with running large-scale, multi-tenant infrastructure on Kubernetes or similar orchestration systems.
Experience with performance-sensitive systems programming and reducing cold-start latency for ephemeral compute.
Track record of contributing to open-source infrastructure or security tooling.
Interest in how AI agents use tools and code execution, and how that shapes the design of safe execution environments.
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $450,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
Thinking Machines Lab will consider for employment qualified applicants with criminal histories in a manner consistent with the requirements of the California Fair Chance Act, the San Francisco Fair Chance Ordinance, and any other applicable state or local fair chance ordinance or law.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're looking for someone to join Data Operations and help researchers get the data they need to train and evaluate our models.
You'll work directly with researchers to understand what they need, figure out how to get it, and own the work through delivery. Sometimes that means finding the right vendor. Sometimes it means digging into a new domain, finding unusual sources of data, or working through a request we haven't seen before.
This is a hands-on role. You'll spend time talking directly with vendors, reviewing data quality, and building processes and systems that make the work easier to repeat and scale.
Our research priorities change quickly, so the work will too. We're looking for someone who is resourceful, flexible, and comfortable figuring things out as they go.
Work with researchers to understand what data they need and turn open-ended requests into a concrete plan.
Own data projects from sourcing and vendor selection through quality review and delivery.
Find the best vendors, experts, and data sources for different types of work.
Build and maintain strong partnerships with data vendors, including international partners.
Review data with researchers and vendors, identify quality issues, and improve the output.
Go deep on new data domains when needed and quickly learn what good looks like.
Build guidelines, processes, and tools that make recurring work easier and more scalable.
Automate repetitive work where it makes sense while staying close to the work that requires judgment.
Work with legal and other teams on practical ways to manage vendor relationships, and adhere to communication guidelines.
Manage multiple projects against research and model timelines and adjust quickly as priorities change.
Minimum Qualifications:
Experience owning operational, data, product, program, or vendor work from start to finish.
Experience working with data vendors ideally in AI or technology.
Strong problem-solving instincts and a willingness to dig in when there isn't an obvious answer.
Comfortable working directly with researchers or technical teams and turning loosely defined needs into action.
Good judgment about quality and a willingness to inspect the work closely rather than rely only on process.
Strong communication skills across different teams, vendors, and cultures.
Comfortable with ambiguity, changing priorities, and tight timelines.
High initiative and willingness to do whatever part of the job is needed to move the work forward.
Able to add useful structure without overcomplicating things.
Preferred qualifications — we encourage you to apply if you meet some but not all of these
Experience with pre-training, post-training, evaluation, annotation, or other AI data types.
Existing relationships with a diverse set of AI or related data vendors.
Experience at an AI company, data company, marketplace, or other fast-moving technology company.
Experience working with international vendors and cross-border operational or compliance issues.
Product, program, operations, or generalist experience in an environment where you had to build the process as you went.
Experience building simple tooling or automation to reduce manual operational work.
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $250,000 - $350,000.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
Thinking Machines Lab will consider for employment qualified applicants with criminal histories in a manner consistent with the requirements of the California Fair Chance Act, the San Francisco Fair Chance Ordinance, and any other applicable state or local fair chance ordinance or law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nWe're looking for someone to join Data Operations and help researchers get the data they need to train and evaluate our models.\n\nYou'll work directly with researchers to understand what they need, figure out how to get it, and own the work through delivery. Sometimes that means finding the right vendor. Sometimes it means digging into a new domain, finding unusual sources of data, or working through a request we haven't seen before.\n\nThis is a hands-on role. You'll spend time talking directly with vendors, reviewing data quality, and building processes and systems that make the work easier to repeat and scale.\n\nOur research priorities change quickly, so the work will too. We're looking for someone who is resourceful, flexible, and comfortable figuring things out as they go.\n\n\n\n\nWHAT YOU'LL DO\n\n - Work with researchers to understand what data they need and turn open-ended requests into a concrete plan.\n\n - Own data projects from sourcing and vendor selection through quality review and delivery.\n\n - Find the best vendors, experts, and data sources for different types of work.\n\n - Build and maintain strong partnerships with data vendors, including international partners.\n\n - Review data with researchers and vendors, identify quality issues, and improve the output.\n\n - Go deep on new data domains when needed and quickly learn what good looks like.\n\n - Build guidelines, processes, and tools that make recurring work easier and more scalable.\n\n - Automate repetitive work where it makes sense while staying close to the work that requires judgment.\n\n - Work with legal and other teams on practical ways to manage vendor relationships, and adhere to communication guidelines.\n\n - Manage multiple projects against research and model timelines and adjust quickly as priorities change.\n \n \n\n\nSKILLS AND QUALIFICATIONS\n\nMinimum Qualifications:\n\n - Experience owning operational, data, product, program, or vendor work from start to finish.\n\n - Experience working with data vendors ideally in AI or technology.\n\n - Strong problem-solving instincts and a willingness to dig in when there isn't an obvious answer.\n\n - Comfortable working directly with researchers or technical teams and turning loosely defined needs into action.\n\n - Good judgment about quality and a willingness to inspect the work closely rather than rely only on process.\n\n - Strong communication skills across different teams, vendors, and cultures.\n\n - Comfortable with ambiguity, changing priorities, and tight timelines.\n\n - High initiative and willingness to do whatever part of the job is needed to move the work forward.\n\n - Able to add useful structure without overcomplicating things.\n\nPreferred qualifications — we encourage you to apply if you meet some but not all of these\n\n - Experience with pre-training, post-training, evaluation, annotation, or other AI data types.\n\n - Existing relationships with a diverse set of AI or related data vendors.\n\n - Experience at an AI company, data company, marketplace, or other fast-moving technology company.\n\n - Experience working with international vendors and cross-border operational or compliance issues.\n\n - Product, program, operations, or generalist experience in an environment where you had to build the process as you went.\n\n - Experience building simple tooling or automation to reduce manual operational work.\n \n \n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, CA.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $250,000 - $350,000.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.\n\nThinking Machines Lab will consider for employment qualified applicants with criminal histories in a manner consistent with the requirements of the California Fair Chance Act, the San Francisco Fair Chance Ordinance, and any other applicable state or local fair chance ordinance or law.\n\n"},{"id":"602f2a99-34eb-4fde-9eec-b4945ee4aab6","title":"Research, Post-Training Evals","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-08-27T02:29:43.309+00:00","isListed":true,"isRemote":true,"workplaceType":"Hybrid","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/602f2a99-34eb-4fde-9eec-b4945ee4aab6","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/602f2a99-34eb-4fde-9eec-b4945ee4aab6/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for a researcher to help develop reliable model evaluations for research signals. This role spans evaluation creation, usability, auditing, and efficiency.
You’ll work closely with researchers and engineers across post-training and the broader research organization. Depending on your interests and experience, you may focus on one area or work across several of these problems.
Create internal evaluations and research signals for capabilities and behaviors important to model research and post-training.
Develop usability evaluations that measure whether models are genuinely useful in real research and product workflows, and partner with the data flywheel to turn evaluation insights into better data and training signals.
Improve evaluation robustness, including grader reliability, ambiguous ground truth, evaluator disagreement, false positives and negatives, and gaps between measured and intended behavior.
Build benchmark auditing methodologies that help researchers understand, trust, and appropriately use evaluation signals.
Develop specialized agentic evaluation environments and user simulators, including supporting harness development and studying cross-user, cross-harness and cross-environment generalization.
Develop evaluations for personalized preferences, biases, values, and other nuanced dimensions of model behavior in collaboration with post-training crafting.
Minimum qualifications:
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Experience designing, building, or analyzing evaluations, benchmarks, datasets, graders, or other measurement systems.
Strong written and verbal communication skills, with the ability to collaborate effectively across research and engineering teams.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
Experience with LLMs, post-training, reinforcement learning, or agentic systems.
Experience with evaluation auditing, human evaluations, LLM-judges, or open-ended task evaluation.
Experience with agentic evaluation, harnesses, long-horizon tasks, or RL environments.
Experience evaluating preferences, personalization, biases, values, or other nuanced model behaviors.
Track record of developing new evaluation methodologies or research signals that meaningfully influenced model development.
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Strong research judgment: clean ablations, honest baselines, and clear technical writing.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
The Coding Agents team makes our models world-class at agentic coding — writing, debugging, and reasoning about code across long-horizon, multi-turn tasks.
You'll join a small, high-leverage team responsible for the recipes, data, and infrastructure behind coding capability gains in every model release.
The team owns the full coding post-training stack: synthetic and human data generation, RL environments and sandboxes, reward and grading design, and large-scale training runs.
This is a research role with real ownership — you'll shape technical direction, not just execute against a spec.
Note: This is an \"evergreen role\" that we keep open on an on-going basis to express interest in this research area. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.
Design and run RL training jobs targeting agentic coding capabilities, iterating on recipes and data.
Build and improve the sandboxed coding environments and reward signals that models are trained and evaluated against.
Generate and curate high-quality synthetic coding data, and build scalable, general-purpose data pipelines.
Design evals that measure real-world coding usefulness, and train models against them to deliver concrete improvements in day-to-day usability.
Debug and analyze large RL runs to catch confounders, reward hacking, and other RL failure modes.
Collaborate closely with infra, evals, and other post-training teams on shared data, joint training runs, and usability improvements — and ship the results into model releases.
Minimum qualifications:
Strong engineering skills, ability to contribute code and debug in complex codebases.
Proficiency in Python and familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX). Comfortable with debugging distributed training and writing code that scales.
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Clarity in communication, an ability to explain complex technical concepts in writing.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but at least some:
Experience building synthetic data pipelines and systems that were adopted by others on your team and remain in use today.
Experience owning the end-to-end cycle of identifying gaps in model usability and closing them through custom evaluations and training data.
Experience making large-scale agentic RL infrastructure reliable given the long tail of failures that surface at scale.
Experience improving the coding capabilities of a frontier model.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Evaluation is one of the most important pillars of building frontier AI systems. It guides research direction, powers experimentation, and helps us understand whether changes to data and training are improving the capabilities and behaviors we care about.
To support this work, researchers need a powerful, self-serve platform that makes it easy to author evaluations, run them or reproduce them reliably at scale, and extract insight from the results. The platform must support both standardized external benchmarks and fast-moving internal evaluations, many kinds of tasks and graders, and inspection from aggregate metrics down to individual model trajectories.
In this role, you will design and build this platform end to end. You will work across Python frameworks, data pipelines, APIs, and user-facing applications, and collaborate closely with pre-training, post-training, and applied teams to improve how we evaluate models and turn results into research decisions.
Design, build, and maintain the platform for authoring, running, tracking, and analyzing model evaluations that are critical in day to day work and model releases.
Work across evaluation libraries, distributed backend systems, data pipelines, APIs, and user-facing applications to deliver capabilities end to end.
Build flexible abstractions for evaluation tasks, environments, graders, datasets, and model outputs without constraining fast-moving research.
Make evaluation results reproducible and trustworthy through versioning, provenance, observability, failure recovery, and robust quality controls.
Partner directly with researchers to identify bottlenecks and turn bespoke workflows into self-serve systems that work across teams.
Work with Research Tooling, the engineering team behind Thinking Machines’ internal research platform.
A bachelor’s degree, or equivalent practical experience, in computer science, engineering, machine learning, or a related field.
Two years of post-grad work experience as a software engineer or ML engineer, exclusive of internships.
Hands-on experience building or maintaining evaluations, benchmarks, graders, or model-quality systems for large language or multimodal models.
Strong software engineering fundamentals and experience building reliable, maintainable systems.
Proficiency in at least one backend programming language; we primarily use Python and Rust. We use React and Typescript on the frontend.
Experience with databases, data pipelines, distributed systems, or other data-intensive infrastructure.
Comfort working across the stack and owning projects from initial problem discovery through deployment and operation.
Experience collaborating with cross-functional partners and subject-matter experts.
We encourage you to apply even if you meet only some of these:
A track record of building frameworks, SDKs, or developer tools with thoughtful abstractions and a strong user experience.
Experience with distributed job execution, workflow orchestration, sandboxed environments, or large-scale data processing.
Experience building polished, intuitive interfaces for inspecting complex data, comparing experiments, or debugging model behavior.
Familiarity with large language or multimodal model evaluation, including model-based grading, human evaluation, or synthetic data.
Experience working closely with researchers to understand their workflows and turn rapidly evolving needs into durable systems.
Experience at a startup or on a small team, building technically complex products end to end.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We are a team of full stack generalists with strong product instincts who work closely with researchers. We build systems that compound research and engineering velocity over time. We own the internal platform researchers use every day to manage and monitor training runs and evaluations, inspect and debug model trajectories, and compare results on shared leaderboards.
You’ll own key parts of this platform, including evaluation and training libraries, experiment-tracking systems, and visualization tools. You’ll identify researchers’ most important bottlenecks and turn them into reliable, generalizable systems. Our team is still small—expect to participate in research meetings, build close relationships with researchers, and gather feedback frequently to develop conviction about where we should invest next.
This role requires technical judgment, close cross-functional collaboration, and product intuition. Success means researchers trust your systems to work, rely on them every day, and find them genuinely delightful to use.
Design, build, and maintain research infra, including evaluation frameworks, training systems, experiment tracking platforms, and visualization tools.
Work across backend systems, data pipelines, and user-facing applications to deliver tools end to end.
Partner directly with researchers to identify bottlenecks and unlock new capabilities. Treat research tooling as a product: proactively gather feedback, set priorities, and measure adoption.
Build systems for reproducibility, traceability, and robust quality control across research experiments and model training runs, with monitoring and observability built in.
A bachelor’s degree, or equivalent practical experience, in computer science, engineering, machine learning, or a related field.
Two years of post-grad work experience as a software engineer or ML engineer, exclusive of internships.
Strong software engineering fundamentals and experience building reliable, maintainable systems.
Proficiency in at least one backend programming language; we primarily use Python and Rust. We use React and Typescript on the frontend.
Experience working with databases, data warehouses (Clickhouse), caching systems such as Redis, and other data infra.
Comfort working across the stack and owning projects from initial problem discovery through deployment and operation.
Experience collaborating with cross-functional partners and subject-matter experts.
We encourage you to apply even if you meet only some of these:
A track record of building tools for researchers, improving developer productivity, or creating technical products for technical users.
Experience building polished, intuitive user-facing applications that demonstrate strong product judgment and attention to detail on UI/UX.
Experience at a startup or on a small team, building technically complex products end to end.
Experience building or maintaining ML research infrastructure, such as training frameworks, evaluation libraries, or experiment-tracking systems.
Experience working closely with researchers to understand and solve their tooling needs.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We're hiring a Full Stack Engineer to help conceive, build, and ship new AI-powered products from the ground up. You'll work on early-stage product ideas that sit at the edge of what our models can do — moving from prototype to something real people rely on, often within weeks rather than quarters.
This role sits at the intersection of product, design, and research. You'll partner closely with researchers to translate emerging model capabilities into interfaces and experiences, and with product and design to figure out what's worth building next. You'll frequently be the first engineer on a given idea, which means real ownership over both the technical direction and the product itself.
We're looking for someone who is energized by ambiguity, comfortable working across the entire stack, and motivated by getting new things in front of real users quickly.
Design, build, and ship full stack features across our web applications — from backend services and APIs to production-quality frontend interfaces.
Prototype rapidly with new and emerging model capabilities, turning early research results into usable product experiences.
Take new product ideas from a rough concept to a live surface that real users can try, iterating quickly based on feedback.
Partner with researchers to understand model behavior, constraints, and capabilities, and translate them into product decisions.
Work with design and product partners to define what to build, not just how to build it.
Own the technical architecture of the products you build, including decisions about scalability, reliability, and maintainability as they mature.
Help set engineering practices and technical standards for a growing product engineering team.
5+ years of professional software engineering experience, with demonstrated ownership across both frontend and backend systems.
Strong track record shipping production web applications end to end, from data model and API design through polished user-facing interfaces.
Comfort working in ambiguous, fast-changing environments where product direction is still being discovered.
Experience making pragmatic technical tradeoffs to move quickly without accumulating unmanageable risk.
Experience building products on top of large language models or other frontier AI systems, including working around their quirks and limitations.
A strong product sense — you enjoy figuring out what's worth building as much as how to build it.
Experience as an early or founding engineer at a startup, or leading a product from 0 to 1.
Familiarity with modern frontend frameworks and a good eye for interface design and usability.
Experience with distributed systems, infrastructure, or backend performance work at scale.
A history of working closely and effectively with researchers, designers, or other non-engineering partners.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
About the Role
This is a product role for someone who can operate inside frontier research without trying to turn research into a conventional software roadmap. The work requires judgment, technical fluency, user empathy, discretion, and the ability to create clarity without creating bureaucracy.
The Post-Training Product Manager will work as a high-trust partner to our post-training researchers. Your role is to understand the work deeply enough to ask the right questions, identify missing connections, surface implications, and help the team decide what matters next. The role sits at the seam between research, model behavior, data and environments, evaluations, training and inference infrastructure, safety, product, and the people using our models.
Inkling was post-trained across math, agentic code and tool use, audio, image, chat, and safety, with a large-scale asynchronous RL program that exceeded 30M rollouts. Inkling and Inkling-Small share a scalable post-training stack, and Inkling is available for customization on Tinker.
The next phase requires tight loops among research priorities, model behavior, evaluations, infrastructure constraints, user and customer learning, and product direction. The person in this role will help those loops compound instead of fragment — maintaining the bird's-eye view while researchers go deep: where work is converging, where teams are solving adjacent problems without enough shared context, what evidence is missing, what decisions are blocked, and how research becomes a stronger model and a useful product.
What You'll Do
Partner with post-training research leaders on priorities, sequencing, decision points, and the connection between research work and the broader model and product agenda
Maintain a clear view across SFT, RL, data and environments, evaluations, safety, model behavior, inference, training infrastructure, and product dependencies; identify gaps before they become blockers
Translate ambiguous research and product questions into concrete learning plans: what must be true, what evidence would change the decision, which experiments or user signals matter, and when the team should revisit the direction
Build lightweight operating mechanisms for critical work — owners, state, dependencies, decisions, risks, release criteria, and follow-through — without imposing a software-development process on research
Bring qualitative and quantitative evidence about model behavior and real workflows into research prioritization, ensuring user knowledge is represented accurately rather than flattened into feature requests
Connect research, product, infrastructure, safety, and leadership when a decision spans teams or when local optimization creates a broader product or model tradeoff
Support the path from research result to usable capability: internal adoption, evaluation, documentation, release readiness, product integration, and feedback after launch
Write clear narratives that explain what the team has learned, what remains uncertain, what decisions are needed, and why the work matters
Take on the unowned work that is necessary to move a critical research-product outcome forward
Skills and Qualifications
Minimum qualifications:
Experience as an early or first product leader in a technical startup, AI lab, research organization, or new product area where the role and operating model were not defined for you
Experience working closely with model training, post-training, RL, evaluations, data, safety, inference, developer platforms, or another technically demanding research-product area
Ability to understand research deeply enough to earn trust, ask sharp questions, and connect technical choices to model behavior and users without overstating your expertise
Strong track record identifying the missing connection, unresolved assumption, or cross-team decision that specialists may not see while deep in the work
Preferred qualifications:
Ability to represent user and product truth in a research environment without reducing research to a list of customer requests
Comfortable making progress when the goal, metric, or path is still evolving, and the correct next step may be a learning loop rather than a launch
Clear communicator who handles disagreement without ego and is willing to change a recommendation when the evidence changes
Motivated by senior IC ownership and proximity to the work more than a large PM team or a conventional product ladder
Background as a research product leader or early AI product leader at a frontier lab, model company, or technically ambitious startup; a technical founder, former engineer, or applied scientist who moved into product; a PM for model training, post-training, evaluation platforms, data systems, ML infrastructure, or developer platforms; or the first PM at a company that turned a novel technical capability into a product, category, or developer ecosystem
Logistics
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $450,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
This role is responsible for building and strengthening the engineering foundations that our post-training research teams depend on. You'll embed within a research team, and build or improve the systems required for the team to succeed. Our research teams are small, and the role carries a corresponding degree of autonomy and responsibility.
Embed within a research team to build, harden, and improve the systems and infrastructure required for the team to succeed.
Design, build, and operate infrastructure research teams depend on, including RL training systems, sandboxing, data pipelines, and agent scaffolding.
Minimum qualifications:
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Experience leading projects end to end, working in large fast-moving codebases, and writing code others have depended and built on.
Strong proficiency in Python and strong engineering fundamentals, with experience debugging systems that fail intermittently and at scale.
Clarity in communication, an ability to explain complex technical concepts in writing.
Strong autonomous drive to progress towards the team’s goals with an ownership mindset.
Preferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but preferably some:
Experience building or iterating with sandboxed or containerized execution environments at a large scale.
Experience in a role where the team's priorities set yours, and a track record of success in such a role.
Familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX).
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nThis role is responsible for building and strengthening the engineering foundations that our post-training research teams depend on. You'll embed within a research team, and build or improve the systems required for the team to succeed. Our research teams are small, and the role carries a corresponding degree of autonomy and responsibility.\n\n\nWHAT YOU’LL DO\n\n - Embed within a research team to build, harden, and improve the systems and infrastructure required for the team to succeed.\n\n - Design, build, and operate infrastructure research teams depend on, including RL training systems, sandboxing, data pipelines, and agent scaffolding.\n\n\nSKILLS AND QUALIFICATIONS\n\nMinimum qualifications:\n\n - Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.\n\n - Experience leading projects end to end, working in large fast-moving codebases, and writing code others have depended and built on.\n\n - Strong proficiency in Python and strong engineering fundamentals, with experience debugging systems that fail intermittently and at scale.\n\n - Clarity in communication, an ability to explain complex technical concepts in writing.\n\n - Strong autonomous drive to progress towards the team’s goals with an ownership mindset.\n\nPreferred qualifications — we encourage you to apply even if you don’t meet all preferred qualifications, but preferably some:\n\n - Experience building or iterating with sandboxed or containerized execution environments at a large scale.\n\n - Experience in a role where the team's priorities set yours, and a track record of success in such a role.\n\n - Familiarity with at least one deep learning framework (e.g., PyTorch, TensorFlow, or JAX).\n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000 - $475,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."},{"id":"919c3fa0-9c77-4770-b839-18492a4dbe6c","title":"Senior Counsel, Commercial & Product","department":"Legal","team":"Legal","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-09-14T16:40:41.722+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/919c3fa0-9c77-4770-b839-18492a4dbe6c","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/919c3fa0-9c77-4770-b839-18492a4dbe6c/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
We’re looking for an attorney to lead the agreements that support our research and products, and support our compliance and counseling for those activities.
You’ll work directly with teams across research, engineering, product, go-to-market, security, safety, privacy, and product counseling to support agreements across a broad range of areas, including enterprise deals, data licensing and tech transactions, grants and research collaborations, product and safety partnerships, and open source considerations. Your work will shape how we partner with vendors, collaborators, and customers to develop and share our technology with the world. You’ll also support our counseling and compliance on our core AI research and products.
Advise research, engineering, product, and cross-functional partners on legal issues arising from agreements and partnerships supporting the development and release of advanced AI.
Draft, negotiate, and review agreements across domains, including enterprise deals, data licensing and tech transactions, grants and research collaborations, product partnerships, safety partnerships, and novel arrangements.
Partner closely with product, research, and go-to-market teams to support company timelines while also supporting core legal strategies.
Own legal infrastructure–including developing playbooks, systems, automation, outside counsel relationships–to support deal velocity, efficiency, and quality.
Support compliance and counseling on research, product, and partnership pipelines, operations, initiatives and launches.
Minimum qualifications:
8+ years of relevant legal experience, with significant expertise in technology transactions and IP and privacy issues.
Client counseling skills, with the ability to exercise judgment under uncertainty and communicate practical, solution-focused advice.
Experience working directly with technical staff and technical counterparties.
Preferred qualifications—we encourage you to apply if you meet some but not all of these:
In-house experience at a technology company shipping products with novel legal risk, ideally on legal issues for AI products, services, or research, with knowledge of the research development lifecycle
Experience negotiating and drafting agreements involving enterprise technology products, data licensing, research collaborations and grant programs, safety partnerships, or product integrations, ideally for AI products, services, or research.
Experience advising on open source licensing.
Strong familiarity with global legal and regulatory frameworks in AI, data privacy, and intellectual property.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000-$360,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nWe’re looking for an attorney to lead the agreements that support our research and products, and support our compliance and counseling for those activities.\n\nYou’ll work directly with teams across research, engineering, product, go-to-market, security, safety, privacy, and product counseling to support agreements across a broad range of areas, including enterprise deals, data licensing and tech transactions, grants and research collaborations, product and safety partnerships, and open source considerations. Your work will shape how we partner with vendors, collaborators, and customers to develop and share our technology with the world. You’ll also support our counseling and compliance on our core AI research and products.\n\n\nWHAT YOU'LL DO\n\n - Advise research, engineering, product, and cross-functional partners on legal issues arising from agreements and partnerships supporting the development and release of advanced AI.\n\n - Draft, negotiate, and review agreements across domains, including enterprise deals, data licensing and tech transactions, grants and research collaborations, product partnerships, safety partnerships, and novel arrangements.\n\n - Partner closely with product, research, and go-to-market teams to support company timelines while also supporting core legal strategies.\n\n - Own legal infrastructure–including developing playbooks, systems, automation, outside counsel relationships–to support deal velocity, efficiency, and quality.\n\n - Support compliance and counseling on research, product, and partnership pipelines, operations, initiatives and launches.\n\n\nSKILLS AND QUALIFICATIONS\n\nMinimum qualifications:\n\n - 8+ years of relevant legal experience, with significant expertise in technology transactions and IP and privacy issues.\n\n - Client counseling skills, with the ability to exercise judgment under uncertainty and communicate practical, solution-focused advice.\n\n - Experience working directly with technical staff and technical counterparties.\n\nPreferred qualifications—we encourage you to apply if you meet some but not all of these:\n\n - In-house experience at a technology company shipping products with novel legal risk, ideally on legal issues for AI products, services, or research, with knowledge of the research development lifecycle\n\n - Experience negotiating and drafting agreements involving enterprise technology products, data licensing, research collaborations and grant programs, safety partnerships, or product integrations, ideally for AI products, services, or research.\n\n - Experience advising on open source licensing.\n\n - Strong familiarity with global legal and regulatory frameworks in AI, data privacy, and intellectual property.\n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $300,000-$360,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\n\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."},{"id":"39d630ce-af8d-404c-a7eb-cbaac2f96b24","title":"Technical Recruiter","department":"Operations","team":"Operations","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-09-16T18:54:40.100+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/39d630ce-af8d-404c-a7eb-cbaac2f96b24","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/39d630ce-af8d-404c-a7eb-cbaac2f96b24/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
About the Role
We're hiring a Technical Recruiter, to lead full-cycle recruiting for the infrastructure eams building and operating the systems behind our frontier model training and Tinker's platform - including cluster and network engineering, distributed systems, and site reliability. You'll partner closely with infrastructure leaders and engineers to understand niche hiring needs and translate them into sourcing strategies, calibrated interview loops, and craft a closing process that lands top candidates in a competitive market.
This is a high-ownership role at a company where infrastructure is the foundation of everything we ship. You'll need to go deep enough on distributed training, networking, and cluster orchestration to evaluate technical signal yourself, not just pattern-match keywords, and to speak credibly with the engineers you're recruiting.
What You’ll Do
Own full-cycle recruiting — sourcing, screening, interview coordination, and closing — for infrastructure, systems, and networking roles.
Partner directly with hiring managers and technical leaders on role scoping and interview loop design
Build and run proactive sourcing strategies and outbound campaigns to reach passive, highly specialized infrastructure talent
Serve as a subject matter expert across distributed training, cluster networking, Kubernetes/container orchestration, and systems programming languages, with the ability to speak to the team's work in depth while assessing candidates
Calibrate talent bar alongside engineering leaders and help interviewers give clear, well-supported feedback
Manage a high-touch candidate experience and act as a trusted advisor to candidates through offer and close
Track pipeline health and hiring metrics, and use that data to continuously improve the hiring process
Help build the recruiting playbooks, tools, and processes that will scale as the infrastructure org grows
Skills & Qualifications
Minimum Qualifications
4+ years of full-cycle technical recruiting experience, with a track record of hiring for infrastructure, platform, or systems engineering roles
Ability to hold a substantive technical conversation about infrastructure domains such as distributed systems, cluster or cloud networking, container orchestration, and systems languages
Experience partnering directly with hiring managers on intake, bar calibration, and interview loop design
Sound independent judgment on candidate quality, and the ability to manage multiple concurrent searches with minimal oversight
Preferred Qualifications
Experience recruiting at a high-growth AI, machine learning, or infrastructure-focused technology company
Proficiency with an applicant tracking system (e.g., Greenhouse or Ashby) and modern sourcing tools
Experience building sourcing strategy or recruiting processes from scratch at an early-stage or fast-scaling team
A genuine interest in Thinking Machines' mission and in the role a world-class infrastructure team plays in achieving it
Logistics
Location: This role is based in San Francisco, CA.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $200,000 - $275,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
Thinking Machines Lab will consider for employment qualified applicants with criminal histories in a manner consistent with the requirements of the California Fair Chance Act, the San Francisco Fair Chance Ordinance, and any other applicable state or local fair chance ordinance or law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\nAbout the Role\n\nWe're hiring a Technical Recruiter, to lead full-cycle recruiting for the infrastructure eams building and operating the systems behind our frontier model training and Tinker's platform - including cluster and network engineering, distributed systems, and site reliability. You'll partner closely with infrastructure leaders and engineers to understand niche hiring needs and translate them into sourcing strategies, calibrated interview loops, and craft a closing process that lands top candidates in a competitive market.\n\nThis is a high-ownership role at a company where infrastructure is the foundation of everything we ship. You'll need to go deep enough on distributed training, networking, and cluster orchestration to evaluate technical signal yourself, not just pattern-match keywords, and to speak credibly with the engineers you're recruiting.\n\nWhat You’ll Do\n\n - Own full-cycle recruiting — sourcing, screening, interview coordination, and closing — for infrastructure, systems, and networking roles.\n\n - Partner directly with hiring managers and technical leaders on role scoping and interview loop design\n\n - Build and run proactive sourcing strategies and outbound campaigns to reach passive, highly specialized infrastructure talent\n\n - Serve as a subject matter expert across distributed training, cluster networking, Kubernetes/container orchestration, and systems programming languages, with the ability to speak to the team's work in depth while assessing candidates\n\n - Calibrate talent bar alongside engineering leaders and help interviewers give clear, well-supported feedback\n\n - Manage a high-touch candidate experience and act as a trusted advisor to candidates through offer and close\n\n - Track pipeline health and hiring metrics, and use that data to continuously improve the hiring process\n\n - Help build the recruiting playbooks, tools, and processes that will scale as the infrastructure org grows\n\nSkills & Qualifications\n\nMinimum Qualifications\n\n - 4+ years of full-cycle technical recruiting experience, with a track record of hiring for infrastructure, platform, or systems engineering roles\n\n - Ability to hold a substantive technical conversation about infrastructure domains such as distributed systems, cluster or cloud networking, container orchestration, and systems languages\n\n - Experience partnering directly with hiring managers on intake, bar calibration, and interview loop design\n\n - Sound independent judgment on candidate quality, and the ability to manage multiple concurrent searches with minimal oversight\n\nPreferred Qualifications\n\n - Experience recruiting at a high-growth AI, machine learning, or infrastructure-focused technology company\n\n - Proficiency with an applicant tracking system (e.g., Greenhouse or Ashby) and modern sourcing tools\n\n - Experience building sourcing strategy or recruiting processes from scratch at an early-stage or fast-scaling team\n\n - A genuine interest in Thinking Machines' mission and in the role a world-class infrastructure team plays in achieving it\n\nLogistics\n\n - Location: This role is based in San Francisco, CA.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $200,000 - $275,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.\n\nThinking Machines Lab will consider for employment qualified applicants with criminal histories in a manner consistent with the requirements of the California Fair Chance Act, the San Francisco Fair Chance Ordinance, and any other applicable state or local fair chance ordinance or law."},{"id":"0e3920c4-9811-491e-9aef-67601c65a87f","title":"Research, Tinker, RL Systems","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-09-16T23:16:40.685+00:00","isListed":true,"isRemote":true,"workplaceType":"Hybrid","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/0e3920c4-9811-491e-9aef-67601c65a87f","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/0e3920c4-9811-491e-9aef-67601c65a87f/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
Tinker is our fine-tuning API that empowers researchers and developers to customize frontier AI to their needs to open access to capabilities that have previously been concentrated in a handful of labs. We manage the infrastructure while allowing Tinkerers full flexibility in training models with their own data, algorithms, and for their own needs.
This role is all about building our training systems for Tinker, including RL systems, numerics, kernels, and beyond.
In this role, you'll develop frontier customization techniques and help build the best post-training engine in the industry, drawing on a whole-stack understanding recipes, data pipelines, and training systems (numerics, kernels, and beyond).
You'll engage directly with the researchers and companies pushing Tinker to its limits. This role is working with both our internal research teams as well as contributing to open science and external partners.
You’ll co-design RL algorithms and training systems across the whole stack, from RL science down to numerics and kernels, to enable anyone to post-train frontier models. You’ll debug RL runs in the wild, optimize post-training pipelines, and help users reach frontier-level results, which in turn makes our platform and models the best they can be.
Required qualifications:
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.
Clarity in communication, an ability to explain complex technical concepts in writing.
Strong interest in working on Tinker and increasing usefulness and adoption.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
Experience with RL training stability techniques for large runs.
Familiarity with low-precision training and inference: numerics, quantization, and their implications for RL.
Hands-on work with LLM serving stacks (e.g., SGLang, vLLM, TokenSpeed, or custom engines).
Experience with scaling studies for large models.
Contributions to open-source training or inference frameworks.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nTinker https://thinkingmachines.ai/tinker/ is our fine-tuning API that empowers researchers and developers to customize frontier AI to their needs to open access to capabilities that have previously been concentrated in a handful of labs. We manage the infrastructure while allowing Tinkerers full flexibility in training models with their own data, algorithms, and for their own needs.\n\nThis role is all about building our training systems for Tinker, including RL systems, numerics, kernels, and beyond.\n\n\n\n\nWHAT YOU’LL DO\n\nIn this role, you'll develop frontier customization techniques and help build the best post-training engine in the industry, drawing on a whole-stack understanding recipes, data pipelines, and training systems (numerics, kernels, and beyond).\n\nYou'll engage directly with the researchers and companies pushing Tinker to its limits. This role is working with both our internal research teams as well as contributing to open science and external partners.\n\nYou’ll co-design RL algorithms and training systems across the whole stack, from RL science down to numerics and kernels, to enable anyone to post-train frontier models. You’ll debug RL runs in the wild, optimize post-training pipelines, and help users reach frontier-level results, which in turn makes our platform and models the best they can be.\n\n\n\n\nSKILLS AND QUALIFICATIONS\n\nRequired qualifications:\n\n - Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.\n\n - Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.\n\n - Clarity in communication, an ability to explain complex technical concepts in writing.\n\n - Strong interest in working on Tinker and increasing usefulness and adoption.\n\nPreferred qualifications — we encourage you to apply if you meet some but not all of these:\n\n - A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.\n\n - Experience with RL training stability techniques for large runs.\n\n - Familiarity with low-precision training and inference: numerics, quantization, and their implications for RL.\n\n - Hands-on work with LLM serving stacks (e.g., SGLang, vLLM, TokenSpeed, or custom engines).\n\n - Experience with scaling studies for large models.\n\n - Contributions to open-source training or inference frameworks.\n\n - PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.\n\n\n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."},{"id":"4e3a366f-eb0f-4ea9-8cf9-9c2f911f279c","title":"Research Lead, Tinker, Fine-tuning Science","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-09-16T23:20:27.811+00:00","isListed":true,"isRemote":true,"workplaceType":"Hybrid","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/4e3a366f-eb0f-4ea9-8cf9-9c2f911f279c","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/4e3a366f-eb0f-4ea9-8cf9-9c2f911f279c/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
At Thinking Machines we build tools that enable people to make AI their own, customizing models to serve their unique needs. This includes the ability to train model weights.
You'll lead the Fine-tuning Science team, setting the research agenda for frontier customization techniques and for Tinker, the leading post-training engine. This is a player-coach role: you'll stay hands-on in the science while growing the team, shaping its direction, and making sure findings ship into Tinker. You'll work closely with our internal research teams, contribute to open science, and engage with external users.
In this role, you'll advance the science of fine-tuning and frontier post-training techniques. You’ll:
Set the research agenda: choose the problems, place the bets, and own the roadmap for pushing Tinker quality, efficiency, and reliability to the frontier.
Lead and grow the team: hire, mentor, and develop researchers, and set the bar for experimental rigor and research taste.
Stay hands-on in areas like LoRA and parameter-efficient fine-tuning and how they interact with RL and post-training.
Improve the stability, efficiency, and reliability of large-scale fine-tuning and RL runs on Tinker.
Represent the work externally through papers, technical blog posts, and community contributions.
Required qualifications:
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
A track record of leading research – setting direction for a team or a major research effort and delivering on it.
Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.
Clarity in communication: an ability to explain complex technical concepts in writing and to build alignment across science, systems/infra, product.
Strong interest in our mission to enable custom models.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
Prior experience with RLHF, RLAIF, preference modeling, or reward learning for large models.
Experience managing or analyzing human data collection campaigns or large-scale annotation workflows.
Research or engineering contributions in alignment, data-centric AI, or human-AI collaboration.
Experience with RL training stability techniques for large runs.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $475,000 - $530,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nAt Thinking Machines we build tools that enable people to make AI their own, customizing models to serve their unique needs. This includes the ability to train model weights.\n\nYou'll lead the Fine-tuning Science team, setting the research agenda for frontier customization techniques and for Tinker, the leading post-training engine. This is a player-coach role: you'll stay hands-on in the science while growing the team, shaping its direction, and making sure findings ship into Tinker. You'll work closely with our internal research teams, contribute to open science, and engage with external users.\n\n \n\n\nWHAT YOU’LL DO\n\n \n\nIn this role, you'll advance the science of fine-tuning and frontier post-training techniques. You’ll:\n\n - Set the research agenda: choose the problems, place the bets, and own the roadmap for pushing Tinker quality, efficiency, and reliability to the frontier.\n\n - Lead and grow the team: hire, mentor, and develop researchers, and set the bar for experimental rigor and research taste.\n\n - Stay hands-on in areas like LoRA and parameter-efficient fine-tuning and how they interact with RL and post-training.\n\n - Improve the stability, efficiency, and reliability of large-scale fine-tuning and RL runs on Tinker.\n\n - Represent the work externally through papers, technical blog posts, and community contributions.\n\n \n \n\n\nSKILLS AND QUALIFICATIONS\n\nRequired qualifications:\n\n - Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.\n\n - A track record of leading research – setting direction for a team or a major research effort and delivering on it.\n\n - Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.\n\n - Clarity in communication: an ability to explain complex technical concepts in writing and to build alignment across science, systems/infra, product.\n\n - Strong interest in our mission to enable custom models.\n\n \n\nPreferred qualifications — we encourage you to apply if you meet some but not all of these:\n\n - A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.\n\n - Prior experience with RLHF, RLAIF, preference modeling, or reward learning for large models.\n\n - Experience managing or analyzing human data collection campaigns or large-scale annotation workflows.\n\n - Research or engineering contributions in alignment, data-centric AI, or human-AI collaboration.\n\n - Experience with RL training stability techniques for large runs.\n\n - PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.\n\n\n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California. \n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $475,000 - $530,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."},{"id":"0a71cb88-068b-4ac6-9912-e498c49b1343","title":"Research, Finetuning Science","department":"Research","team":"Research","employmentType":"FullTime","location":"San Francisco","secondaryLocations":[],"publishedAt":"2026-09-16T23:21:35.061+00:00","isListed":true,"isRemote":true,"workplaceType":"Hybrid","address":{"postalAddress":{"postalCode":"94110","addressRegion":"California","addressCountry":"United States","addressLocality":"San Francisco"}},"jobUrl":"https://jobs.ashbyhq.com/thinkingmachines/0a71cb88-068b-4ac6-9912-e498c49b1343","applyUrl":"https://jobs.ashbyhq.com/thinkingmachines/0a71cb88-068b-4ac6-9912-e498c49b1343/application","descriptionHtml":"The mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.
At Thinking Machines we build tools that enable people to make AI their own, customizing models to serve their unique needs. This includes the ability to train model weights.
In this role, you'll work on frontier customization techniques and help build the best post-training engine in the industry – Tinker – drawing on a whole-stack understanding of RL science. Findings directly shape Tinker's training defaults, API design, and the open-source Tinker Cookbook. You'll work with our internal research teams as well as contributing to open science for external partners.
In this role, you'll advance the science of fine-tuning and frontier post-training techniques. You’ll:
Contribute to areas like LoRA and parameter efficient fine-tuning and how to push customization quality, efficiency, and reliability to the frontier.
Ship research into product: inform Tinker's training defaults and primitives, and codify best-practice methods as recipes in the Tinker Cookbook.
Improve the stability, efficiency, and reliability of large-scale fine-tuning and RL runs on Tinker.
Share what you learn through papers, technical blog posts, and community contributions.
You’ll contribute to areas like LoRA, parameter-efficient fine-tuning, how things interact with RL and post-training, and how to push customization quality, efficiency, and reliability to the frontier.
Required qualifications:
Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.
Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.
Clarity in communication, an ability to explain complex technical concepts in writing.
Strong interest in our mission to enable custom models.
Preferred qualifications — we encourage you to apply if you meet some but not all of these:
A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.
Prior experience with RLHF, RLAIF, preference modeling, or reward learning for large models.
Experience managing or analyzing human data collection campaigns or large-scale annotation workflows.
Research or engineering contributions in alignment, data-centric AI, or human-AI collaboration.
Experience with RL training stability techniques for large runs.
PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.
Location: This role is based in San Francisco, California.
Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.
Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.
Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.
As set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law.
","descriptionPlain":"ABOUT THINKING MACHINES\n\nThe mission of Thinking Machines is to build AI that extends human will and judgment. We are training frontier models with Inkling, developing Tinker to let people make models their own, and crafting interfaces that broaden human-AI communication. We believe the future worth building is human, and we're hiring people who want to build it.\n\n\n\n\nABOUT THE ROLE\n\nAt Thinking Machines we build tools that enable people to make AI their own, customizing models to serve their unique needs. This includes the ability to train model weights.\n\nIn this role, you'll work on frontier customization techniques and help build the best post-training engine in the industry – Tinker – drawing on a whole-stack understanding of RL science. Findings directly shape Tinker's training defaults, API design, and the open-source Tinker Cookbook. You'll work with our internal research teams as well as contributing to open science for external partners.\n\n \n\n\nWHAT YOU’LL DO\n\n \n\nIn this role, you'll advance the science of fine-tuning and frontier post-training techniques. You’ll:\n\n - Contribute to areas like LoRA and parameter efficient fine-tuning and how to push customization quality, efficiency, and reliability to the frontier.\n\n - Ship research into product: inform Tinker's training defaults and primitives, and codify best-practice methods as recipes in the Tinker Cookbook.\n\n - Improve the stability, efficiency, and reliability of large-scale fine-tuning and RL runs on Tinker.\n\n - Share what you learn through papers, technical blog posts, and community contributions.\n\nYou’ll contribute to areas like LoRA, parameter-efficient fine-tuning, how things interact with RL and post-training, and how to push customization quality, efficiency, and reliability to the frontier.\n\n \n\n\nSKILLS AND QUALIFICATIONS\n\nRequired qualifications:\n\n - Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding.\n\n - Proficiency in Python and familiarity with deep learning frameworks (e.g., PyTorch, TensorFlow, or JAX). Comfort debugging distributed training and writing code that scales.\n\n - Clarity in communication, an ability to explain complex technical concepts in writing.\n\n - Strong interest in our mission to enable custom models.\n\nPreferred qualifications — we encourage you to apply if you meet some but not all of these:\n\n - A strong grasp of probability, statistics, and ML fundamentals. You can look at experimental data and distinguish between real effects, noise, and bugs.\n\n - Prior experience with RLHF, RLAIF, preference modeling, or reward learning for large models.\n\n - Experience managing or analyzing human data collection campaigns or large-scale annotation workflows.\n\n - Research or engineering contributions in alignment, data-centric AI, or human-AI collaboration.\n\n - Experience with RL training stability techniques for large runs.\n\n - PhD in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline with strong theoretical and empirical grounding; or, equivalent industry research experience.\n\n\nLOGISTICS\n\n - Location: This role is based in San Francisco, California.\n\n - Compensation: Depending on background, skills and experience, the expected annual salary range for this position is $350,000 - $475,000 USD.\n\n - Visa sponsorship: We sponsor visas. While we can't guarantee success for every candidate or role, if you're the right fit, we're committed to working through the visa process together.\n\n - Benefits: Thinking Machines offers generous health, dental, and vision benefits, unlimited PTO, paid parental leave, and relocation support as needed.\n\nAs set forth in Thinking Machines' Equal Employment Opportunity policy, we do not discriminate on the basis of any protected group status under any applicable law."}],"apiVersion":"1"}