At Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.
Inherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.
We’re looking for Members of Technical Staff to lead work on post-training state-of-the-art foundation models for open-ended agentic capabilities in scientific research. You’ll be involved at every level of the post-training pipeline: sourcing and creating data, building autocurricula, devising and implementing SFT and RL algorithms, constructing tools and harnesses for foundation model self-improvement, analysing research results, and using information gained to devise future hypotheses. You will work closely with an experienced technical team of humans, and increasingly alongside the AI scientist collaborators we dogfood.
Design, implement, and tune SFT and RL algorithms to post-train models that autonomously perform state-of-the-art research.
Build the autocurricula, judges, harnesses and eval pipelines that turn open-ended research tasks into reliable reward signal.
Run large-scale experiments on state-of-the-art hardware and analyse experiments to determine the next hypotheses to test, in collaboration with our AI agents.
Close recursive loops so that AI agents drive their own post-training research.
Work closely with colleagues in the Infrastructure and AI for Science teams to optimise hardware and deliver remarkable performance in real scientific domains.
3+ years of deep learning research experience.
Experience post-training large language, vision, video or multi-modal models.
Demonstrated track record of success in deep learning research, whether papers, model releases, open-source contributions, or other artifacts.
5+ years of software engineering experience, including deep familiarity with Python and at least one deep learning framework (e.g., PyTorch, JAX).
Experience using the latest coding agents, and opinions about optimal workflow.
Enthusiasm for experimental organizational design.
AI-pilled: adopting agents, keen to build a company where agents are front and centre.
PhD in mathematics, computer science or hard science discipline.
Hands-on experience training LLMs with RL at scale (GRPO/PPO, DPO, distillation, and variants).
Familiarity with distributed and long-context training infrastructure.
A background in autocurricula, open-endedness, meta-learning, or recursive self-improvement.
Experience post-training frontier models at an industry lab (scale, infra, and iteration speed).
You'll shape the core research of a frontier AI lab from the beginning.
You'll work on genuine recursive self-improvement — training AI scientists that improve the very pipeline that trains them — not incremental benchmark-chasing.
You'll dogfood your own work: the agents you post-train accelerate the research that creates them.
Small team, high trust, no bureaucracy, and a genuinely technical culture.
If you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation.
","descriptionPlain":"MEMBER OF TECHNICAL STAFF, POST-TRAINING — INHERENT (LONDON)\n\nAt Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.\n\nInherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.\n\n\n\n\nABOUT THE ROLE\n\nWe’re looking for Members of Technical Staff to lead work on post-training state-of-the-art foundation models for open-ended agentic capabilities in scientific research. You’ll be involved at every level of the post-training pipeline: sourcing and creating data, building autocurricula, devising and implementing SFT and RL algorithms, constructing tools and harnesses for foundation model self-improvement, analysing research results, and using information gained to devise future hypotheses. You will work closely with an experienced technical team of humans, and increasingly alongside the AI scientist collaborators we dogfood.\n\n\n\n\nWHAT YOU'D DO\n\n - Design, implement, and tune SFT and RL algorithms to post-train models that autonomously perform state-of-the-art research.\n\n - Build the autocurricula, judges, harnesses and eval pipelines that turn open-ended research tasks into reliable reward signal.\n\n - Run large-scale experiments on state-of-the-art hardware and analyse experiments to determine the next hypotheses to test, in collaboration with our AI agents.\n\n - Close recursive loops so that AI agents drive their own post-training research.\n\n - Work closely with colleagues in the Infrastructure and AI for Science teams to optimise hardware and deliver remarkable performance in real scientific domains.\n \n \n\n\nWHAT WE'RE LOOKING FOR\n\n - 3+ years of deep learning research experience.\n\n - Experience post-training large language, vision, video or multi-modal models.\n\n - Demonstrated track record of success in deep learning research, whether papers, model releases, open-source contributions, or other artifacts.\n\n - 5+ years of software engineering experience, including deep familiarity with Python and at least one deep learning framework (e.g., PyTorch, JAX).\n\n - Experience using the latest coding agents, and opinions about optimal workflow.\n\n - Enthusiasm for experimental organizational design.\n\n - AI-pilled: adopting agents, keen to build a company where agents are front and centre.\n \n \n\n\nSTRONG CANDIDATES MAY ALSO HAVE\n\n - PhD in mathematics, computer science or hard science discipline.\n\n - Hands-on experience training LLMs with RL at scale (GRPO/PPO, DPO, distillation, and variants).\n\n - Familiarity with distributed and long-context training infrastructure.\n\n - A background in autocurricula, open-endedness, meta-learning, or recursive self-improvement.\n\n - Experience post-training frontier models at an industry lab (scale, infra, and iteration speed).\n \n \n\n\nWHY THIS IS INTERESTING\n\n - You'll shape the core research of a frontier AI lab from the beginning.\n\n - You'll work on genuine recursive self-improvement — training AI scientists that improve the very pipeline that trains them — not incremental benchmark-chasing.\n\n - You'll dogfood your own work: the agents you post-train accelerate the research that creates them.\n\n - Small team, high trust, no bureaucracy, and a genuinely technical culture.\n \n \n\n\nCULTURE\n\nIf you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation."},{"id":"2013628f-ec56-4244-84f2-cbcd85a1fe42","title":"Member of Technical Staff (AI for Science)","department":"Technical Staff","team":"Technical Staff","employmentType":"FullTime","location":"London","secondaryLocations":[],"publishedAt":"2026-05-28T17:26:03.341+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}},"jobUrl":"https://jobs.ashbyhq.com/inherent/2013628f-ec56-4244-84f2-cbcd85a1fe42","applyUrl":"https://jobs.ashbyhq.com/inherent/2013628f-ec56-4244-84f2-cbcd85a1fe42/application","descriptionHtml":"At Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.
Inherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.
We’re looking for Members of Technical Staff to develop transformative AI Scientist agents that make meaningful contributions to in silico and in situ discoveries in specific scientific domains. Your work will leverage our state-of-the-art proprietary foundation models, adapting their capabilities to a particular scientific vertical. You will be involved in every-level of the AI for Science lifecycle: judiciously selecting research problems, creating evaluations and training data, post-training and harnessing Inherent’s models, building interfaces to infrastructure for scientific experiments (e.g., scientific software, simulators, cloud labs), driving experimental iteration on agent behaviour, and collaborating with external scientists to validate and deploy our systems. You will work closely with an experienced technical team of humans, and increasingly alongside the AI scientist collaborators we dogfood.
Devise and hone AI for Science research settings in your area of expertise that are amenable to open-ended exploration by AI Scientist agents (either in silico or in situ).
Benchmark Inherent’s proprietary agents in your AI for Science environments, gathering data to improve agent performance via post-training and harness iteration.
Contributing data, evaluations, feedback and expert opinion to Inherent’s post-training team, in service of improving our core foundation models.
Close recursive loops wherever possible, for example by enabling our agents to automatically deploy, debug and repair themselves.
Work closely with scientific partners and customers, which may include forward deployment.
5+ years of experience in hard science research, AI for science, data science or industry R&D.
Experience applying ML to drive scientific progress in silico or in situ on in-lab or real-world datasets.
Demonstrated track record of success in research, whether papers, product releases, open-source contributions, or other artifacts.
3+ years of software engineering experience, including familiarity with Python and at least one deep learning framework.
Experience using the latest coding agents, and opinions about optimal workflow.
Enthusiasm for experimental organizational design.
AI-pilled: adopting agents, keen to build a company where agents are front and centre.
PhD in a scientific discipline.
Contribution to a well-cited work in AI for Science writ large (e.g., ML to predict protein structure, neural weather forecasting systems, reinforcement learning for materials discovery, or causal models for financial risk prediction).
Hands-on experience using AI agents to automate parts of the scientific workflow in a particular scientific vertical.
Familiarity with post-training and harnessing foundation models.
An interest in AI Scientist agents, open-endedness, meta-learning, or recursive self-improvement.
Strong relationships with leading scientists in a field of scientific research amenable to low-latency and high-throughput experiments.
Experience building or operating autonomous labs.
You'll shape the core research of a frontier AI lab from the beginning.
You'll work on genuine scientific discovery — training AI scientists that empower humans to flourish — not incremental benchmark-chasing.
Small team, high trust, no bureaucracy, and a genuinely technical culture.
Our products will be used by the world’s best scientists to power breakthrough research of huge benefit to humanity, while preserving human agency.
If you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation.
","descriptionPlain":"MEMBER OF TECHNICAL STAFF, AI FOR SCIENCE — INHERENT (LONDON)\n\nAt Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.\n\nInherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.\n\n\n\n\nABOUT THE ROLE\n\nWe’re looking for Members of Technical Staff to develop transformative AI Scientist agents that make meaningful contributions to in silico and in situ discoveries in specific scientific domains. Your work will leverage our state-of-the-art proprietary foundation models, adapting their capabilities to a particular scientific vertical. You will be involved in every-level of the AI for Science lifecycle: judiciously selecting research problems, creating evaluations and training data, post-training and harnessing Inherent’s models, building interfaces to infrastructure for scientific experiments (e.g., scientific software, simulators, cloud labs), driving experimental iteration on agent behaviour, and collaborating with external scientists to validate and deploy our systems. You will work closely with an experienced technical team of humans, and increasingly alongside the AI scientist collaborators we dogfood.\n\n\n\n\nWHAT YOU'D DO\n\n - Devise and hone AI for Science research settings in your area of expertise that are amenable to open-ended exploration by AI Scientist agents (either in silico or in situ).\n\n - Benchmark Inherent’s proprietary agents in your AI for Science environments, gathering data to improve agent performance via post-training and harness iteration.\n\n - Contributing data, evaluations, feedback and expert opinion to Inherent’s post-training team, in service of improving our core foundation models.\n\n - Close recursive loops wherever possible, for example by enabling our agents to automatically deploy, debug and repair themselves.\n\n - Work closely with scientific partners and customers, which may include forward deployment.\n\n\n\n\nWHAT WE'RE LOOKING FOR\n\n - 5+ years of experience in hard science research, AI for science, data science or industry R&D.\n\n - Experience applying ML to drive scientific progress in silico or in situ on in-lab or real-world datasets.\n\n - Demonstrated track record of success in research, whether papers, product releases, open-source contributions, or other artifacts.\n\n - 3+ years of software engineering experience, including familiarity with Python and at least one deep learning framework.\n\n - Experience using the latest coding agents, and opinions about optimal workflow.\n\n - Enthusiasm for experimental organizational design.\n\n - AI-pilled: adopting agents, keen to build a company where agents are front and centre.\n \n \n\n\nSTRONG CANDIDATES MAY ALSO HAVE\n\n - PhD in a scientific discipline.\n\n - Contribution to a well-cited work in AI for Science writ large (e.g., ML to predict protein structure, neural weather forecasting systems, reinforcement learning for materials discovery, or causal models for financial risk prediction).\n\n - Hands-on experience using AI agents to automate parts of the scientific workflow in a particular scientific vertical.\n\n - Familiarity with post-training and harnessing foundation models.\n\n - An interest in AI Scientist agents, open-endedness, meta-learning, or recursive self-improvement.\n\n - Strong relationships with leading scientists in a field of scientific research amenable to low-latency and high-throughput experiments.\n\n - Experience building or operating autonomous labs.\n \n \n\n\nWHY THIS IS INTERESTING\n\n - You'll shape the core research of a frontier AI lab from the beginning.\n\n - You'll work on genuine scientific discovery — training AI scientists that empower humans to flourish — not incremental benchmark-chasing.\n\n - Small team, high trust, no bureaucracy, and a genuinely technical culture.\n\n - Our products will be used by the world’s best scientists to power breakthrough research of huge benefit to humanity, while preserving human agency.\n \n \n\n\nCULTURE\n\nIf you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation."},{"id":"bca10c78-8efd-4fd2-bd8c-f9e20b63201e","title":"Member of Technical Staff (Factory Redesign)","department":"Technical Staff","team":"Technical Staff","employmentType":"FullTime","location":"London","secondaryLocations":[],"publishedAt":"2026-05-28T17:27:09.488+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}},"jobUrl":"https://jobs.ashbyhq.com/inherent/bca10c78-8efd-4fd2-bd8c-f9e20b63201e","applyUrl":"https://jobs.ashbyhq.com/inherent/bca10c78-8efd-4fd2-bd8c-f9e20b63201e/application","descriptionHtml":"At Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.
Inherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.
We're looking for someone to help redesign our research factory by inventing new systems, workflows, and measurements for human-machine teaming. You will continually figure out the most effective way to embed self-improving AI within the organisation, to the benefit of human safety, agency, fairness and enjoyability.
You'll define and build novel tooling that centers AI in every part of the company: research, operations, decision-making, planning, hiring. You'll also run rigorous experiments to predict and test empirically how well the new organisational structures work: defining metrics, designing interventions, and surfacing the best human-machine configuration for every kind of work we do.
Factory design is at the heart of how we do research. Inherent is a recursive company through and through: from the infrastructure level, to the scientific level, to the organisational level — and an improvement in each area triggers improvement in the others.
Build novel internal tooling that integrates AI deeply into research workflows, operations, and decision-making.
Design and run experiments measuring the economic and productivity impact of AI across the company.
Invent new metrics for human-machine collaboration — what's working, what isn't, and where the leverage is.
Push the frontier of agentic coding and AI-augmented work internally, inventing new agentic coding methods and workflows.
Partner with every team to find the optimal human-machine configuration for the work they do.
Deep software engineering or data science experience with strong expertise in Python.
Strong fluency with frontier agentic coding tools, e.g. Claude Code, Cursor, Codex — used daily, with opinions on what good workflow looks like.
Comfort designing and analysing experiments: strong analytical skills, rigorous measurement, and experience in experiment design.
Interest in the future of work and how AI should be woven into an organisation.
Good taste: you know when to build, when to buy, and when to delete.
AI-pilled: adopting agents, keen to build a company where agents are front and centre.
A track record of building tooling (internal or external) that other people have adopted.
Background in productivity research, applied economics, or organisational design.
Comfort moving fluidly between writing code and doing quantitative analysis.
Experience instrumenting and measuring engineering or research orgs end-to-end.
You'll shape how a frontier AI lab is run from the beginning, and be an architect of its organisational operating system for years to come.
You’ll devise and participate in radical organisational experiments that help to define the future of work at Inherent and beyond.
Small team, high trust, no bureaucracy, and a genuinely technical culture.
If you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation.
","descriptionPlain":"MEMBER OF TECHNICAL STAFF, FACTORY REDESIGN — INHERENT (LONDON)\n\nAt Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.\n\nInherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.\n\n\n\n\nABOUT THE ROLE\n\nWe're looking for someone to help redesign our research factory by inventing new systems, workflows, and measurements for human-machine teaming. You will continually figure out the most effective way to embed self-improving AI within the organisation, to the benefit of human safety, agency, fairness and enjoyability.\n\nYou'll define and build novel tooling that centers AI in every part of the company: research, operations, decision-making, planning, hiring. You'll also run rigorous experiments to predict and test empirically how well the new organisational structures work: defining metrics, designing interventions, and surfacing the best human-machine configuration for every kind of work we do.\n\nFactory design is at the heart of how we do research. Inherent is a recursive company through and through: from the infrastructure level, to the scientific level, to the organisational level — and an improvement in each area triggers improvement in the others.\n\n\n\n\nWHAT YOU'D DO\n\n - Build novel internal tooling that integrates AI deeply into research workflows, operations, and decision-making.\n\n - Design and run experiments measuring the economic and productivity impact of AI across the company.\n\n - Invent new metrics for human-machine collaboration — what's working, what isn't, and where the leverage is.\n\n - Push the frontier of agentic coding and AI-augmented work internally, inventing new agentic coding methods and workflows.\n\n - Partner with every team to find the optimal human-machine configuration for the work they do.\n \n \n\n\nWHAT WE'RE LOOKING FOR\n\n - Deep software engineering or data science experience with strong expertise in Python.\n\n - Strong fluency with frontier agentic coding tools, e.g. Claude Code, Cursor, Codex — used daily, with opinions on what good workflow looks like.\n\n - Comfort designing and analysing experiments: strong analytical skills, rigorous measurement, and experience in experiment design.\n\n - Interest in the future of work and how AI should be woven into an organisation.\n\n - Good taste: you know when to build, when to buy, and when to delete.\n\n - AI-pilled: adopting agents, keen to build a company where agents are front and centre.\n \n \n\n\nSTRONG CANDIDATES MAY ALSO HAVE\n\n - A track record of building tooling (internal or external) that other people have adopted.\n\n - Background in productivity research, applied economics, or organisational design.\n\n - Comfort moving fluidly between writing code and doing quantitative analysis.\n\n - Experience instrumenting and measuring engineering or research orgs end-to-end.\n \n \n\n\nWHY THIS IS INTERESTING\n\n - You'll shape how a frontier AI lab is run from the beginning, and be an architect of its organisational operating system for years to come.\n\n - You’ll devise and participate in radical organisational experiments that help to define the future of work at Inherent and beyond.\n\n - Small team, high trust, no bureaucracy, and a genuinely technical culture.\n \n \n\n\nCULTURE\n\nIf you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation."},{"id":"a5f6a788-677e-423f-9245-edb1a87f1125","title":"Member of Technical Staff (Infrastructure Engineer, Training and Inference Systems)","department":"Technical Staff","team":"Technical Staff","employmentType":"FullTime","location":"London","secondaryLocations":[],"publishedAt":"2026-05-28T17:46:31.797+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}},"jobUrl":"https://jobs.ashbyhq.com/inherent/a5f6a788-677e-423f-9245-edb1a87f1125","applyUrl":"https://jobs.ashbyhq.com/inherent/a5f6a788-677e-423f-9245-edb1a87f1125/application","descriptionHtml":"At Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.
Inherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.
We're looking for an infrastructure engineer to help build the training and inference systems that frontier research depends on. You'll build and maintain the systems our researchers — and our AI Scientists themselves — depend on to train and serve models. The focus is on speed, reliability, and ease-of-use, with a particular emphasis on large-scale distributed RL and deployment in scientific domains. Infra here is a core part of the research process, not a support function. Inherent is a recursive company through and through, and we’re constantly closing loops from the infra level, to the scientific level, to the org level.
Contribute to our training + inference infrastructure end-to-end.
Build systems to run large-scale experiments: distributed job orchestration, data pipelines, evals.
Implement and harden RL and post-training libraries for our models.
Close recursive loops so AI agents can drive their own training and infrastructure.
Work closely with the core research team on the systems they live inside.
High-performance, large-scale distributed systems experience, preferably for LLM workloads.
Proficiency in Python and PyTorch or JAX.
Hands-on experience with large-scale LLM training or inference technologies, e.g. SGLang, vLLM, verl, Megatron, OpenRLHF.
Good taste: you know when to build, when to buy, and when to delete.
AI-pilled: adopting agents, keen to build a company where agents are front and centre.
Experience with a systems programming language like Rust or C++.
Experience with writing and profiling CUDA kernels.
A track record of building reliable research tools.
You'll shape the core technical foundation of a frontier AI lab from the beginning.
The infra problems are unusually hard and creative: iteration speed for recursively self-improving agents, which themselves compound the iteration speed, is the whole game.
Small team, high trust, no bureaucracy, and a genuinely technical culture.
You’ll work alongside world-class colleagues with diverse backgrounds: experts in foundation model training, AI for science, and organisational design.
If you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation.
","descriptionPlain":"MEMBER OF TECHNICAL STAFF, TRAINING INFRASTRUCTURE — INHERENT (LONDON)\n\nAt Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.\n\nInherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.\n\n\n\n\nABOUT THE ROLE\n\nWe're looking for an infrastructure engineer to help build the training and inference systems that frontier research depends on. You'll build and maintain the systems our researchers — and our AI Scientists themselves — depend on to train and serve models. The focus is on speed, reliability, and ease-of-use, with a particular emphasis on large-scale distributed RL and deployment in scientific domains. Infra here is a core part of the research process, not a support function. Inherent is a recursive company through and through, and we’re constantly closing loops from the infra level, to the scientific level, to the org level.\n\n\n\n\nWHAT YOU'D DO\n\n - Contribute to our training + inference infrastructure end-to-end.\n\n - Build systems to run large-scale experiments: distributed job orchestration, data pipelines, evals.\n\n - Implement and harden RL and post-training libraries for our models.\n\n - Close recursive loops so AI agents can drive their own training and infrastructure.\n\n - Work closely with the core research team on the systems they live inside.\n \n \n\n\nWHAT WE'RE LOOKING FOR\n\n - High-performance, large-scale distributed systems experience, preferably for LLM workloads.\n\n - Proficiency in Python and PyTorch or JAX.\n\n - Hands-on experience with large-scale LLM training or inference technologies, e.g. SGLang, vLLM, verl, Megatron, OpenRLHF.\n\n - Good taste: you know when to build, when to buy, and when to delete.\n\n - AI-pilled: adopting agents, keen to build a company where agents are front and centre.\n \n \n\n\nSTRONG CANDIDATES MAY ALSO HAVE\n\n - Experience with a systems programming language like Rust or C++.\n\n - Experience with writing and profiling CUDA kernels.\n\n - A track record of building reliable research tools.\n \n \n\n\nWHY THIS IS INTERESTING\n\n - You'll shape the core technical foundation of a frontier AI lab from the beginning.\n\n - The infra problems are unusually hard and creative: iteration speed for recursively self-improving agents, which themselves compound the iteration speed, is the whole game.\n\n - Small team, high trust, no bureaucracy, and a genuinely technical culture.\n\n - You’ll work alongside world-class colleagues with diverse backgrounds: experts in foundation model training, AI for science, and organisational design.\n \n \n\n\nCULTURE\n\nIf you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation."},{"id":"32ab1a96-660d-4c42-9ff6-e8e8b22e9551","title":"Member of Technical Staff (Infrastructure Engineer, Compute Infrastructure)","department":"Technical Staff","team":"Technical Staff","employmentType":"FullTime","location":"London","secondaryLocations":[],"publishedAt":"2026-05-28T17:27:18.532+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}},"jobUrl":"https://jobs.ashbyhq.com/inherent/32ab1a96-660d-4c42-9ff6-e8e8b22e9551","applyUrl":"https://jobs.ashbyhq.com/inherent/32ab1a96-660d-4c42-9ff6-e8e8b22e9551/application","descriptionHtml":"At Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.
Inherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.
We're looking for an infrastructure engineer to help make best use of cutting-edge hardware for inference and training. You'll build the operational layer of our research compute: GPU clusters, scheduling, networking, observability, and the on-call system that keeps it all running. This is the foundation that lets frontier research happen quickly, reliably, and repeatedly for both humans and AI agents. Infra here is a core part of the research process, not a support function. Inherent is a recursive company through and through, and we’re constantly closing loops from the infra level, to the scientific level, to the org level.
Run and evolve our GPU clusters: scheduling, utilisation, debugging, performance.
Scale the Kubernetes / Linux / networking / cloud stack end-to-end.
Establish operational excellence: incident response, postmortem culture, on-call health.
Build agent-driven automation for cluster lifecycle, provisioning, and remediation.
Partner directly with the research team to optimise the platform they rely on every day.
Experience operating infra at scale, preferably for LLM workloads.
Depth in Kubernetes internals, cluster provisioning, and orchestration systems.
Comfortable across the stack: Kubernetes, Linux, networking, containers, cloud environments.
Strong systems thinking: you care about reliability, performance, and operational clarity.
Good taste: you know when to build, when to buy, and when to delete.
AI-pilled: adopting agents, keen to build a company where agents are front and centre.
Cloud and cluster networking expertise, e.g. VPC, BGP, CNI, eBPF, service mesh.
Experience with GPUs and CUDA.
Infrastructure-as-code and workflow orchestration experience (Terraform and similar).
Track record leading multi-quarter infra initiatives end-to-end.
You'll shape the core technical foundation of a frontier AI lab from the beginning.
The infra problems are unusually hard and creative: iteration speed for recursively self-improving agents, which themselves compound the iteration speed, is the whole game.
Small team, high trust, no bureaucracy, and a genuinely technical culture.
You’ll work alongside world-class colleagues with diverse backgrounds: experts in foundation model training, AI for science, and organisational design.
If you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation.
","descriptionPlain":"MEMBER OF TECHNICAL STAFF, COMPUTE INFRASTRUCTURE — INHERENT (LONDON)\n\nAt Inherent, we are on a mission to build AI that recursively self-improves to discover new knowledge. Scientific advances are the backbone of our economic, technological and societal prosperity, but ideas are getting harder to find and breakthroughs are becoming more expensive. We are building a new frontier lab dedicated to developing AI that explores “unknown unknowns” to uncover paradigm-shifting research contributions. Science is a social endeavour, and so our mission is inextricably a human-machine teaming problem. We’re starting by reinventing the AI research factory so that our own agents accelerate their own creation.\n\nInherent is a well-funded, fast-growing neo-lab backed by Tier 1 VCs who believe in our ethical stance. We are a team of operators with backgrounds at frontier labs who have done foundational work in recursive self-improvement, AI Scientists, world modelling, meta-RL and human-machine cooperation. Working in-person every day at our high-intensity London headquarters, we believe that Europe will lead the way in the coming paradigm of AI-enabled science, unlocking human potential across the globe.\n\n\n\n\nABOUT THE ROLE\n\nWe're looking for an infrastructure engineer to help make best use of cutting-edge hardware for inference and training. You'll build the operational layer of our research compute: GPU clusters, scheduling, networking, observability, and the on-call system that keeps it all running. This is the foundation that lets frontier research happen quickly, reliably, and repeatedly for both humans and AI agents. Infra here is a core part of the research process, not a support function. Inherent is a recursive company through and through, and we’re constantly closing loops from the infra level, to the scientific level, to the org level.\n\n\n\n\nWHAT YOU'D DO\n\n - Run and evolve our GPU clusters: scheduling, utilisation, debugging, performance.\n\n - Scale the Kubernetes / Linux / networking / cloud stack end-to-end.\n\n - Establish operational excellence: incident response, postmortem culture, on-call health.\n\n - Build agent-driven automation for cluster lifecycle, provisioning, and remediation.\n\n - Partner directly with the research team to optimise the platform they rely on every day.\n \n \n\n\nWHAT WE'RE LOOKING FOR\n\n - Experience operating infra at scale, preferably for LLM workloads.\n\n - Depth in Kubernetes internals, cluster provisioning, and orchestration systems.\n\n - Comfortable across the stack: Kubernetes, Linux, networking, containers, cloud environments.\n\n - Strong systems thinking: you care about reliability, performance, and operational clarity.\n\n - Good taste: you know when to build, when to buy, and when to delete.\n\n - AI-pilled: adopting agents, keen to build a company where agents are front and centre.\n \n \n\n\nSTRONG CANDIDATES MAY ALSO HAVE\n\n - Cloud and cluster networking expertise, e.g. VPC, BGP, CNI, eBPF, service mesh.\n\n - Experience with GPUs and CUDA.\n\n - Infrastructure-as-code and workflow orchestration experience (Terraform and similar).\n\n - Track record leading multi-quarter infra initiatives end-to-end.\n \n \n\n\nWHY THIS IS INTERESTING\n\n - You'll shape the core technical foundation of a frontier AI lab from the beginning.\n\n - The infra problems are unusually hard and creative: iteration speed for recursively self-improving agents, which themselves compound the iteration speed, is the whole game.\n\n - Small team, high trust, no bureaucracy, and a genuinely technical culture.\n\n - You’ll work alongside world-class colleagues with diverse backgrounds: experts in foundation model training, AI for science, and organisational design.\n \n \n\n\nCULTURE\n\nIf you believe in our mission and culture, and are qualified and motivated, we encourage you to apply, even if you don’t meet every one of the criteria above. We know that many of the most creative and talented people have had unusual career paths and backgrounds. Building a team with a diversity of thought is mission-critical, for plurality spurs curiosity, invention and collective experimentation."},{"id":"159a9b86-71e1-4b6e-b408-e5aae55bc568","title":"Talent Architect ","department":"Operations Staff","team":"Operations Staff","employmentType":"FullTime","location":"London","secondaryLocations":[],"publishedAt":"2026-10-14T10:44:51.344+00:00","isListed":true,"isRemote":false,"workplaceType":"OnSite","address":{"postalAddress":{"addressRegion":"England","addressCountry":"United Kingdom","addressLocality":"London"}},"jobUrl":"https://jobs.ashbyhq.com/inherent/159a9b86-71e1-4b6e-b408-e5aae55bc568","applyUrl":"https://jobs.ashbyhq.com/inherent/159a9b86-71e1-4b6e-b408-e5aae55bc568/application","descriptionHtml":"King’s Cross, London, with international travel · £110,000–£130,000 (flex for the right candidate)
Almost every leap in how we live, from medicine to energy to computing, began with a scientific breakthrough. But breakthroughs are getting harder to find and more expensive to make. We think AI can change that.
As our Talent Architect, you’ll assemble the team that builds Inherent: focusing on high impact researchers and engineers. You’ll do it with total access to our founding team, and you’ll help invent an AI-native way of doing it.
What we’re building
Today’s AI is brilliant at giving good answers. It’s far worse at working out which questions are worth asking, and it still can’t learn on the fly. Yet many of the discoveries that changed the world, like penicillin, X-rays and the microwave oven, came from stumbling into unknown unknowns.
Inherent exists to build AI that can do that: AI that recursively improves itself to discover new knowledge, working alongside human scientists on the hardest problems we face.
Our core bet is that bolting AI onto old ways of working only brings small gains. Electricity didn’t transform industry until factories were redesigned around it, and we think the research lab is no different. So we’re rebuilding ours from scratch. Faraday, our AI scientist, works as a full member of the team. It reads papers, joins meetings, runs experiments, manages compute and even helps with hiring. As Faraday improves, the whole company speeds up, and how humans and agents work together is itself part of our research.
We’re doing this with care. Inherent is a public benefit corporation, and we think hard about the second- and third-order consequences of what we build.
Current state of play
Our founders have been at the frontier of this work for a decade. We’ve brought together a small team of elite researchers, and applications are growing every week.
This year we raised a $50m seed round led by Tier 1 VCs, and we’re growing fast. The next people through the door will decide what this lab becomes, and you’ll be the one who finds them.
What you’ll do
Fundamentally your job is to become one of the most well connected people in frontier AI and scientific R&D, with the freedom and responsibility to build a truly new kind of organization.
Find people before the market does. Map talent across post-training, AI for science and infrastructure, looking hardest for people others have overlooked.
Build an elite network. Find and win over exceptional researchers and engineers from top academic and industry labs, the people who rarely look at job ads.
Represent Inherent around the world. Travel to conferences such as NeurIPS, ICML and ICLR, meet researchers in person, and help run our own dinners, seminars and hackathons.
Make everyone a talent architect. Build and track referrals across the team, compounding our unique culture of technical excellence, deep commitment to societal benefit, and creative weirdness.
Keep us fast. Run a pipeline and interview experience that’s quick, thoughtful and that candidates remember for the right reasons.
Centre the machine. Work with Faraday, our internal agent, and in close collaboration with the “Reinventing the Factory” team to turn hiring into a system that improves itself, so every search, screen and conversation makes the next one better.
Why you’ll want this
Few people early in their careers get to build an AI-native talent function from scratch, working side by side with founders. Here, that’s the job. Every system you build makes you faster, so your reach compounds month on month, much like the agents we’re building. You’ll learn from the people at the frontier of the field, with a path to wider operational leadership as we scale.
What you’ll bring
A genuine fascination with the future of scientific discovery and why AI could change it.
Curiosity about AI research: you read papers, follow the labs and want to understand the work.
New ideas about how to build a world-leading team and its associated brand: we want you to build architecture for talent differently, not to follow a generic playbook.
Good judgement about people, and a willingness to learn to spot exceptional researchers from a CV, a GitHub profile or a paper trail.
A passion for uncovering fascinating people: you enjoy finding the person whom few others might have noticed.
Clear, warm, distinctive writing. Outreach that gets replies because it’s specific, credible and unusual.
Organisation to keep many conversations moving without dropping a thread.
Ability to coordinate distinctive, tasteful events that reflect Inherent’s intellectual ethos.
Discretion with sensitive information about candidates, pay and company plans.
You already use AI to multiply your time, and want to push that much further.
A desire to participate in and build a culture that values weirdness, creativity, artistic excellence, scientific precision, hard work and kindness.
It helps if you’ve done some recruiting or sourcing, studied or worked in a scientific or technical field, or built a community around something you care about. But none of this is essential. We’d rather hire someone sharp and mission-driven early than someone experienced and indifferent.
Worth knowing
We’re in person every day in King’s Cross, the pace is high, and there’s no playbook for hiring. You’ll travel regularly for conferences and events. If you want a well-oiled process to run, this isn’t it. If you want to build one, it is.
Apply
If that sounds like you, apply, even if you don’t tick every box. Some of the best people we know took strange routes to get here.
","descriptionPlain":"King’s Cross, London, with international travel · £110,000–£130,000 (flex for the right candidate)\n\nAlmost every leap in how we live, from medicine to energy to computing, began with a scientific breakthrough. But breakthroughs are getting harder to find and more expensive to make. We think AI can change that.\n\nAs our Talent Architect, you’ll assemble the team that builds Inherent: focusing on high impact researchers and engineers. You’ll do it with total access to our founding team, and you’ll help invent an AI-native way of doing it.\n\nWhat we’re building\n\nToday’s AI is brilliant at giving good answers. It’s far worse at working out which questions are worth asking, and it still can’t learn on the fly. Yet many of the discoveries that changed the world, like penicillin, X-rays and the microwave oven, came from stumbling into unknown unknowns.\n\nInherent exists to build AI that can do that: AI that recursively improves itself to discover new knowledge, working alongside human scientists on the hardest problems we face.\n\nOur core bet is that bolting AI onto old ways of working only brings small gains. Electricity didn’t transform industry until factories were redesigned around it, and we think the research lab is no different. So we’re rebuilding ours from scratch. Faraday, our AI scientist, works as a full member of the team. It reads papers, joins meetings, runs experiments, manages compute and even helps with hiring. As Faraday improves, the whole company speeds up, and how humans and agents work together is itself part of our research.\n\nWe’re doing this with care. Inherent is a public benefit corporation, and we think hard about the second- and third-order consequences of what we build.\n\nCurrent state of play\n\nOur founders have been at the frontier of this work for a decade. We’ve brought together a small team of elite researchers, and applications are growing every week.\n\nThis year we raised a $50m seed round led by Tier 1 VCs, and we’re growing fast. The next people through the door will decide what this lab becomes, and you’ll be the one who finds them.\n\nWhat you’ll do\n\nFundamentally your job is to become one of the most well connected people in frontier AI and scientific R&D, with the freedom and responsibility to build a truly new kind of organization.\n\n - Find people before the market does. Map talent across post-training, AI for science and infrastructure, looking hardest for people others have overlooked.\n\n - Build an elite network. Find and win over exceptional researchers and engineers from top academic and industry labs, the people who rarely look at job ads.\n\n - Represent Inherent around the world. Travel to conferences such as NeurIPS, ICML and ICLR, meet researchers in person, and help run our own dinners, seminars and hackathons.\n\n - Make everyone a talent architect. Build and track referrals across the team, compounding our unique culture of technical excellence, deep commitment to societal benefit, and creative weirdness.\n\n - Keep us fast. Run a pipeline and interview experience that’s quick, thoughtful and that candidates remember for the right reasons.\n\n - Centre the machine. Work with Faraday, our internal agent, and in close collaboration with the “Reinventing the Factory” team to turn hiring into a system that improves itself, so every search, screen and conversation makes the next one better.\n\nWhy you’ll want this\n\nFew people early in their careers get to build an AI-native talent function from scratch, working side by side with founders. Here, that’s the job. Every system you build makes you faster, so your reach compounds month on month, much like the agents we’re building. You’ll learn from the people at the frontier of the field, with a path to wider operational leadership as we scale.\n\nWhat you’ll bring\n\n - A genuine fascination with the future of scientific discovery and why AI could change it.\n\n - Curiosity about AI research: you read papers, follow the labs and want to understand the work.\n\n - New ideas about how to build a world-leading team and its associated brand: we want you to build architecture for talent differently, not to follow a generic playbook.\n\n - Good judgement about people, and a willingness to learn to spot exceptional researchers from a CV, a GitHub profile or a paper trail.\n\n - A passion for uncovering fascinating people: you enjoy finding the person whom few others might have noticed.\n\n - Clear, warm, distinctive writing. Outreach that gets replies because it’s specific, credible and unusual.\n\n - Organisation to keep many conversations moving without dropping a thread.\n\n - Ability to coordinate distinctive, tasteful events that reflect Inherent’s intellectual ethos.\n\n - Discretion with sensitive information about candidates, pay and company plans.\n\n - You already use AI to multiply your time, and want to push that much further.\n\n - A desire to participate in and build a culture that values weirdness, creativity, artistic excellence, scientific precision, hard work and kindness.\n\nIt helps if you’ve done some recruiting or sourcing, studied or worked in a scientific or technical field, or built a community around something you care about. But none of this is essential. We’d rather hire someone sharp and mission-driven early than someone experienced and indifferent.\n\nWorth knowing\n\nWe’re in person every day in King’s Cross, the pace is high, and there’s no playbook for hiring. You’ll travel regularly for conferences and events. If you want a well-oiled process to run, this isn’t it. If you want to build one, it is.\n\nApply\n\nIf that sounds like you, apply, even if you don’t tick every box. Some of the best people we know took strange routes to get here."}],"apiVersion":"1"}