Fleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
Fleet moves quickly because a small team has access to reliable systems. As Fleet's first dedicated IT Engineer, you will be responsible for the corporate systems behind that work: identity and access, business applications, employee devices, and office networks.
This is a hands-on, high-ownership role based in Fleet's San Francisco office with on-call responsibility for business-critical systems. You will make technical and vendor decisions, resolve difficult incidents, and build automation so recurring work scales without adding manual overhead.
Manage access. Keep identity, permissions, and provisioning coherent across Fleet's systems.
Make employee transitions boring. Accounts, access, and devices should be ready when needed and removed cleanly when they are not.
Keep corporate systems reliable. Manage employee devices, office networks, and shared technology while resolving incidents and eliminating recurring failures.
Automate the function. Use agents, APIs, code, and workflows to turn repetitive work into controlled, observable, and recoverable systems.
Manage vendors and tooling. Handle licensing, utilization, escalations, and decisions to expand, replace, or retire tools.
Experience running corporate IT, enterprise systems, or a major technology domain end to end in a fast-growing startup.
Comfortable with APIs, scripting, workflow automation, and AI tools as part of how you work.
Hands-on experience with identity and access systems, productivity platforms, and business-critical applications.
Judgment about when a change is worth making, what it will affect, and how to introduce it without unnecessary disruption.
Experience maintaining employee endpoints, device management, and office networks.
Security experience is a plus, especially across identity, endpoints, device trust, or incident response.
$200K - $240K base salary and equity
Salary range: $200,000–$240,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## About the Role\n\nFleet moves quickly because a small team has access to reliable systems. As Fleet's first dedicated IT Engineer, you will be responsible for the corporate systems behind that work: identity and access, business applications, employee devices, and office networks.\n\nThis is a hands-on, high-ownership role based in Fleet's San Francisco office with on-call responsibility for business-critical systems. You will make technical and vendor decisions, resolve difficult incidents, and build automation so recurring work scales without adding manual overhead.\n\n## What You'll Do\n\n- **Manage access.** Keep identity, permissions, and provisioning coherent across Fleet's systems.\n- **Make employee transitions boring.** Accounts, access, and devices should be ready when needed and removed cleanly when they are not.\n- **Keep corporate systems reliable.** Manage employee devices, office networks, and shared technology while resolving incidents and eliminating recurring failures.\n- **Automate the function.** Use agents, APIs, code, and workflows to turn repetitive work into controlled, observable, and recoverable systems.\n- **Manage vendors and tooling.** Handle licensing, utilization, escalations, and decisions to expand, replace, or retire tools.\n\n## What We're Looking For\n\n- Experience running corporate IT, enterprise systems, or a major technology domain end to end in a fast-growing startup.\n- Comfortable with APIs, scripting, workflow automation, and AI tools as part of how you work.\n- Hands-on experience with identity and access systems, productivity platforms, and business-critical applications.\n- Judgment about when a change is worth making, what it will affect, and how to introduce it without unnecessary disruption.\n- Experience maintaining employee endpoints, device management, and office networks.\n- Security experience is a plus, especially across identity, endpoints, device trust, or incident response.\n\n## Compensation\n\nSalary range: $200,000–$240,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-08-18T23:20:50.804307+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"646241db-355a-4205-808b-6a3e5776e816","slug":"former-founder-with-track-record-and-experience","title":"Deployments","subtitle":"Member of Technical Staff","locations":["San Francisco","New York"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"We work with frontier labs, hyperscalers, and enterprises to help develop and deploy the next generation of AI agents. We see the creation of evals and environments, codifying human goals for agents, as the highest leverage human activity in the build up to ASI.
We've raised from Sequoia Capital, Bain Capital Ventures, Menlo Ventures, and SV Angel, are growing very quickly, and are off to the races to build the core infrastructure to unlock double digit economic growth with the next generation of agents.
As Member of Technical Staff, Deployments you will collaborate closely with our entirely technical team to deliver robust environments, realistic data, and challenging tasks that are just of out of reach of today's frontier models.
Many of the projects will require herculean effort, combined with high attention to detail, and research taste. We want people who will do whatever it takes to solve immediate bottlenecks but also think in systems and are not content with local maxima.
You will not be a good fit if you:
Thrive in routine, consistency, or structure, or you aren't good at context switching between tasks or priorities.
Alternatively, we are looking for people who are excited to work on the frontier which is (almost by definition) filled with ambiguity and chaos. Clarity of thought and communication is a must-have.
Seek to work in silos, manage, strategize, and plan without accountability for execution
Alternatively, we are looking for people who aren't afraid to get their hands dirty and are excited to do a bit of everything across the entire tech stack, and across engineering and non-engineering work, including interacting with researchers and leaders at frontier labs
Don't use AI, or aren't excited to spend a lot of time working with and improving AI agents
Alternatively, we are looking for people great at managing many AI agents in parallel and understanding their capabilities, without outsourcing their own taste and judgment that will shape frontier model performance.
Get things done for the sake of getting them done, or are fast and scrappy without care or focus on detail.
Alternatively, we are looking for systems thinkers who don't sacrifice quality for speed; they deeply care about the work they put out, and therefore are incredibly fast but build systems that enable all of us to scale quality.
Experiencing founding, working early-stage, knowledge of AI evals are pluses.
We are on-site in San Francisco and New York. Compensation (salary + equity) will be extremely competitive.
Being an engineer at an early-stage startup is not easy. Here are some of our values and what we care about. If this excites you, we would love to chat.
Own their outcomes: We succeed when our partners succeed. Therefore, we listen, build trust, and deliver value beyond what is contracted. We develop novel environments to study AI safety and capabilities. We post-train our own agents to share our learnings and best-practices. We study model behavior intensely and frequently report observations in our initial evaluation runs.
Relentlessness: We are relentless in the speed at which we move when we build conviction, in getting in front of the right people, and in improving ourselves as a company and as individuals.
Truth over comfort: Integrity to us means being honest with our partners, each other, and most importantly, ourselves. At the end of the day, early-stage startups are about the relentless pursuit of the truth, and we hold ourselves to a high standard to ensure we live that and succeed.
Precision under pressure: We care deeply about delivering infrastructure to train safe, reliable agents. As such, a single detail in an environment can be the reason for unintentional failures within major training runs. Everyone on the team is fast and scrappy, but pays attention to every small detail. We don't cut corners; we cut scope and build systems
Grit: A startup is built by people who like to get their hands dirty. We want to see that when left to your own devices, you know what to do and can stay resilient in the face of uncertainty and ambiguity.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"### Introduction\n\nWe work with frontier labs, hyperscalers, and enterprises to help develop and deploy the next generation of AI agents. We see the creation of evals and environments, codifying human goals for agents, as the highest leverage human activity in the build up to ASI.\n\nWe've raised from Sequoia Capital, Bain Capital Ventures, Menlo Ventures, and SV Angel, are growing very quickly, and are off to the races to build the core infrastructure to unlock double digit economic growth with the next generation of agents.\n\n### **A Brief Description of the Role**\n\nAs Member of Technical Staff, Deployments you will collaborate closely with our entirely technical team to deliver robust environments, realistic data, and challenging tasks that are just of out of reach of today's frontier models.\n\nMany of the projects will require herculean effort, combined with high attention to detail, and research taste. We want people who will do whatever it takes to solve immediate bottlenecks but also think in systems and are not content with local maxima.\n\nYou will not be a good fit if you:\n\n- Thrive in routine, consistency, or structure, or you aren't good at context switching between tasks or priorities.\n - Alternatively, we are looking for people who are excited to work on the frontier which is (almost by definition) filled with ambiguity and chaos. Clarity of thought and communication is a must-have.\n- Seek to work in silos, manage, strategize, and plan without accountability for execution\n - Alternatively, we are looking for people who aren't afraid to get their hands dirty and are excited to do a bit of everything across the entire tech stack, and across engineering and non-engineering work, including interacting with researchers and leaders at frontier labs\n- Don't use AI, or aren't excited to spend a lot of time working with and improving AI agents\n - Alternatively, we are looking for people great at managing many AI agents in parallel and understanding their capabilities, without outsourcing their own taste and judgment that will shape frontier model performance.\n- Get things done for the sake of getting them done, or are fast and scrappy without care or focus on detail.\n - Alternatively, we are looking for systems thinkers who don't sacrifice quality for speed; they deeply care about the work they put out, and therefore are incredibly fast but build systems that enable all of us to scale quality.\n\nExperiencing founding, working early-stage, knowledge of AI evals are pluses.\n\nWe are on-site in San Francisco and New York.\n\n### A Brief Description of Us\n\nBeing an engineer at an early-stage startup is not easy. Here are some of our values and what we care about. If this excites you, we would love to chat.\n\n- **Own their outcomes**: We succeed when our partners succeed. Therefore, we listen, build trust, and deliver value _beyond_ what is contracted. We develop novel environments to study AI safety and capabilities. We post-train our own agents to share our learnings and best-practices. We study model behavior intensely and frequently report observations in our initial evaluation runs.\n- **Relentlessness**: We are relentless in the speed at which we move when we build conviction, in getting in front of the right people, and in improving ourselves as a company and as individuals.\n- **Truth over comfort**: Integrity to us means being honest with our partners, each other, and most importantly, ourselves. At the end of the day, early-stage startups are about the relentless pursuit of the truth, and we hold ourselves to a high standard to ensure we live that and succeed.\n- **Precision under pressure**: We care deeply about delivering infrastructure to train safe, reliable agents. As such, a single detail in an environment can be the reason for unintentional failures within major training runs. Everyone on the team is fast and scrappy, but pays attention to every small detail. We don't cut corners; we cut scope and build systems\n- **Grit**: A startup is built by people who like to get their hands dirty. We want to see that when left to your own devices, you know what to do and can stay resilient in the face of uncertainty and ambiguity.\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-02-19T06:24:11.408829+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"5898a625-6b5f-4369-a944-bb72721d7a95","slug":"research-engineering","title":"Research Engineer, Infrastructure","subtitle":"Member of Technical Staff","locations":["San Francisco"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Fleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
Fleet runs a large GPU cluster. As a Research Engineer, you own the infrastructure our research runs on: the training stack, the inference stack, the research codebase that ties them together, the cluster itself, and the kernels at the bottom. Your job is to keep all of it fast and reliable as the research moves.
You work close to the researchers and fix what is actually slow, where the fix belongs. Sometimes that is a Triton kernel. Sometimes it is a parallelism config change. Sometimes it is deleting a library nobody is using anymore.
The research codebase is under active development and most of the bottlenecks are not obvious from outside it. We want someone who can profile a stalled collective, write a fused kernel, and refactor research code without breaking the science. This is hire #1 for the role.
. On-site in San Francisco.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## About the Role\n\nFleet runs a large GPU cluster. As a Research Engineer, you own the infrastructure our research runs on: the training stack, the inference stack, the research codebase that ties them together, the cluster itself, and the kernels at the bottom. Your job is to keep all of it fast and reliable as the research moves.\n\nYou work close to the researchers and fix what is actually slow, where the fix belongs. Sometimes that is a Triton kernel. Sometimes it is a parallelism config change. Sometimes it is deleting a library nobody is using anymore.\n\nThe research codebase is under active development and most of the bottlenecks are not obvious from outside it. We want someone who can profile a stalled collective, write a fused kernel, and refactor research code without breaking the science. This is hire #1 for the role.\n\n## What You'll Do\n\n- Own the cluster. Capacity, failover, fault tolerance, and on-call when training stalls.\n- Drive utilization. Profile training runs, find the bottleneck (NCCL, KV cache, dataloader, allgather), and fix it. Write kernels when the profiler points there.\n- Maintain training and inference. Distributed training (FSDP, TP, PP), rollout generation, async training, and the inference stack the training loop depends on.\n- Maintain the research codebase. Keep it fast and reliable as researchers extend it. Refactor without breaking the science.\n\n## How We Work\n\n- Small, technical team.\n- Speed with rigor.\n- Truth over comfort.\n- On-site in SF.\n\n## What We're Looking For\n\n- Distributed training depth. Hands-on experience with multi-node training (tens of GPUs at minimum, ideally hundreds) and the major parallelism strategies (DP, TP, PP, FSDP).\n- Performance and kernels. You can profile a slow training run and land the fix. You write CUDA and Triton, and have shipped a fused op or attention variant in a real training run.\n- Inference systems. You have built or maintained a high-throughput inference stack (vLLM, SGLang). KV cache, continuous batching, speculative decoding, and quantization are tools you have used in production.\n- Research-codebase taste. You have walked into a sprawling, half-documented research repo and made it measurably faster or more reliable without breaking the science. RL post-training methods (PPO, GRPO, RLHF) and their failure modes are familiar enough that you can triage a training failure from a log line.\n\n## Especially Exciting\n\n- Contributions to PyTorch, vLLM, SGLang, Megatron, DeepSpeed, FSDP, NCCL, flash-attention, or other open-source ML systems.\n- Experience operating GPU fleets at scale.\n- Experience with RoCE / InfiniBand, GPUDirect, or similar high-performance networking.\n- Prior work on RL rollout systems, async training, or off-policy training infra.\n\n### Compensation\n\nOn-site in San Francisco.\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-06-26T20:23:02.967127+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"918ad0f6-83a8-4c73-8366-27ec3f3087c4","slug":"mts-generalist","title":"Platform","subtitle":"Member of Technical Staff","locations":["San Francisco","New York"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Fleet AI is an applied artificial intelligence company working on human-AI collaboration at scale, backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel. Compute can be bought and text can be scraped. But agents only learn real jobs by doing them, which takes a world where the job exists and the outcome can be graded. That is what we build. Our training gyms are high-fidelity simulations of how real businesses actually run. Agents practice genuine work inside them while people supervise, evaluate and steer their behavior. Each one is a step toward an allocation economy, a future where humans direct work rather than perform it.
Platform is where Fleet’s leverage compounds. Every environment we ship and every training run we grade sits on systems this team owns. A week saved here is a week saved for every engineer and researcher in the company, on every environment, from then on. When the platform is good, one person ships what used to take a team.
The team is accountable for the agent runtime and the execution layer every environment runs on. It also owns the durable record of every run and the production infrastructure the rest of engineering depends on.
Own the agent runtime. Make a long agent rollout a first-class object rather than a process. It should pause, resume at the exact step it stopped, and stay steerable while it is still running.
Make environments ship in one motion. Carry an environment from a repository to a graded deployment along a single dependable path, so that shipping the hundredth one feels like shipping the first.
Design the systems of record. Give every task, trajectory and verifier one durable home. The contract should be clear enough that other teams can build against it without asking you first.
Make long jobs boring. Multi-hour work that checkpoints cleanly, retries honestly, and can tell you exactly where it is at any moment.
Put numbers on model and harness choices. Build the evaluation that shows what a change does to quality, latency and cost, so those tradeoffs get argued from data.
Turn one team’s solution into everyone’s capability. When something works well for Research or Environment Engineering or Deployments, make it a platform primitive the whole company inherits.
Strong backend and systems fundamentals, including production experience keeping something distributed or asynchronous alive
Range across infrastructure and interface, from designing a queue one day to building the internal tool for inspecting it the next
Judgment about when a shared abstraction earns its keep and when a targeted fix is the more honest answer
A habit of instrumenting first, so you can say what a change did to reliability, to cost, or to somebody’s afternoon
Comfort chasing a failure across service boundaries and carrying the fix all the way into production
Fluency with AI coding agents as part of how you work
Full-time and on-site in San Francisco or New York, with an extremely competitive salary and meaningful equity. Apply even if you don’t match every line above. How you build and think matters more to us than credentials.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"Fleet AI is an applied artificial intelligence company working on human-AI collaboration at scale, backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel. Compute can be bought and text can be scraped. But agents only learn real jobs by doing them, which takes a world where the job exists and the outcome can be graded. That is what we build. Our training gyms are high-fidelity simulations of how real businesses actually run. Agents practice genuine work inside them while people supervise, evaluate and steer their behavior. Each one is a step toward an allocation economy, a future where humans direct work rather than perform it.\n\n### **About the Role**\n\nPlatform is where Fleet’s leverage compounds. Every environment we ship and every training run we grade sits on systems this team owns. A week saved here is a week saved for every engineer and researcher in the company, on every environment, from then on. When the platform is good, one person ships what used to take a team.\n\nThe team is accountable for the agent runtime and the execution layer every environment runs on. It also owns the durable record of every run and the production infrastructure the rest of engineering depends on.\n\n## What You’ll Do\n\n- **Own the agent runtime.** Make a long agent rollout a first-class object rather than a process. It should pause, resume at the exact step it stopped, and stay steerable while it is still running.\n- **Make environments ship in one motion.** Carry an environment from a repository to a graded deployment along a single dependable path, so that shipping the hundredth one feels like shipping the first.\n- **Design the systems of record.** Give every task, trajectory and verifier one durable home. The contract should be clear enough that other teams can build against it without asking you first.\n- **Make long jobs boring.** Multi-hour work that checkpoints cleanly, retries honestly, and can tell you exactly where it is at any moment.\n- **Put numbers on model and harness choices.** Build the evaluation that shows what a change does to quality, latency and cost, so those tradeoffs get argued from data.\n- **Turn one team’s solution into everyone’s capability.** When something works well for Research or Environment Engineering or Deployments, make it a platform primitive the whole company inherits.\n\n## What We’re Looking For\n\n- Strong backend and systems fundamentals, including production experience keeping something distributed or asynchronous alive\n- Range across infrastructure and interface, from designing a queue one day to building the internal tool for inspecting it the next\n- Judgment about when a shared abstraction earns its keep and when a targeted fix is the more honest answer\n- A habit of instrumenting first, so you can say what a change did to reliability, to cost, or to somebody’s afternoon\n- Comfort chasing a failure across service boundaries and carrying the fix all the way into production\n- Fluency with AI coding agents as part of how you work\n\nFull-time and on-site in San Francisco or New York. Apply even if you don’t match every line above. How you build and think matters more to us than credentials.\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-02-19T06:24:11.645971+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"88527cf1-813f-48dc-afc2-d83c68d86561","slug":"mts-applied-research","title":"Applied Research Scientist","subtitle":"Member of Technical Staff","locations":["San Francisco","New York"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"We work with frontier labs, hyperscalers, and enterprises to build the environments, tasks, and evaluations that the next generation of agents are trained and judged on. We see this work, codifying human goals for agents and exposing where frontier models break, as the highest-leverage activity in the build-up to ASI. We've raised from Sequoia Capital, Menlo Ventures, BCV, and SV Angel, are growing very quickly, and are building the core infrastructure to unlock the next generation of agents.
As an Applied Researcher, you will collaborate closely with our technical team to design and deliver robust environments, realistic data, and challenging tasks that are just out of reach of today's frontier models.
This role is also the bridge between our partner engagements and our core research team. You go deep on a lab engagement, come back with what you saw, and feed that into core research. When you find a thread worth pursuing, you pursue it.
Many projects will require herculean effort, attention to detail, and research taste. We want people who will do whatever it takes to solve immediate bottlenecks, and also think in systems and refuse local maxima. You will be at the frontier of AI development every day, on problems that, if solved, are among the biggest unlocks in front of us.
We want people who can demonstrate they will work hard, learn fast, and go deep in at least one of: applied AI research, low-level infrastructure, or AI agent engineering. A former-founder track record, early-engineer-at-an-early-stage-startup background, or prior AI research experience is a strong plus.
.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet\n\nWe work with frontier labs, hyperscalers, and enterprises to build the environments, tasks, and evaluations that the next generation of agents are trained and judged on. We see this work, codifying human goals for agents and exposing where frontier models break, as the highest-leverage activity in the build-up to ASI. We've raised from Sequoia Capital, Menlo Ventures, BCV, and SV Angel, are growing very quickly, and are building the core infrastructure to unlock the next generation of agents.\n\n## About the Role\n\nAs an Applied Researcher, you will collaborate closely with our technical team to design and deliver robust environments, realistic data, and challenging tasks that are just out of reach of today's frontier models.\n\nThis role is also the bridge between our partner engagements and our core research team. You go deep on a lab engagement, come back with what you saw, and feed that into core research. When you find a thread worth pursuing, you pursue it.\n\nMany projects will require herculean effort, attention to detail, and research taste. We want people who will do whatever it takes to solve immediate bottlenecks, and also think in systems and refuse local maxima. You will be at the frontier of AI development every day, on problems that, if solved, are among the biggest unlocks in front of us.\n\n## What You'll Do\n\n- Understand the ask. Take a lab request and figure out what they actually need: not the literal spec, the underlying capability they're trying to teach or measure. Push back when the ask is wrong.\n- Ship environments, tasks, and data. Deliver an artifact that runs end-to-end and is ready to go into a training or eval run.\n- Be diligent. Know your failure modes. Your artifact will be used to train or evaluate a frontier model. A broken reward, a leaked answer, a degenerate solution path, a flaky environment can corrupt a multi-million-dollar run or invalidate a benchmark. Be the person who knows exactly how their artifact can break, and who has already closed the obvious ways.\n- Close the loop. Each engagement teaches you something about how the lab's model fails, what reward shape actually trained, where an eval held up or didn't. That signal goes back to core research and shapes the next thread, recipe, or env primitive we build. Pursue your own threads when you find one worth pulling.\n- Compound the platform. Reusable harnesses, eval protocols, dataset patterns, and infra primitives you build for one delivery become the substrate for the next.\n\n## What We're Looking For\n\nWe want people who can demonstrate they will work hard, learn fast, and go deep in at least one of: applied AI research, low-level infrastructure, or AI agent engineering. A former-founder track record, early-engineer-at-an-early-stage-startup background, or prior AI research experience is a strong plus.\n\n- Strong engineer. Production Python; comfortable across containers, distributed systems, and a GPU when you have to.\n- Deep understanding of ML. You know how models actually learn from environments: what makes a reward trainable, how RL training dynamics behave in practice, what supervision, preference, and RL data each do, where capability actually comes from.\n- Research taste. You can look at a partner's request and tell whether the environment, the reward, or the data is what's going to determine whether the model learns. You can defend that view to a lab researcher.\n- High ownership. You are the person who ships, and the person who owns the engagement end-to-end.\n- Comfort with ambiguity. The frontier is, almost by definition, chaos. You should be energized by that, not paralyzed.\n- Bonus: depth in a domain we want to teach a model.\n\n## How We Work\n\n- Small, technical team.\n- Speed with rigor.\n- Truth over comfort. Honest with partners, with each other, and with ourselves.\n- On-site in SF or NYC.\n\n### Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-06-17T06:00:44.071063+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"aeaccf6e-eb76-4b54-a758-e79382b462d7","slug":"fleet-fellow-mech-e","title":"Mechanical Engineering Specialist — Contract Engagement","subtitle":null,"locations":["Remote"],"work_type":"Remote","employment_type":"Contract","pipeline_type":"contractor","salary_range":"$60–$100/hr initially; up to $150/hr","equity_range":null,"referral_bonus":500,"body":"Fleet AI builds training environments where AI agents practice real tasks and their work can be evaluated. Our mechanical engineering projects use realistic engineering problems to test how well those agents reason, use tools and produce valid results.
Develop mechanical engineering problems and reference materials for AI training and evaluation. Depending on the project, the work may cover CAD, structural or thermal analysis, fluid flow, manufacturing drawings, or design validation.
Each project will specify the outputs to be delivered. These may include:
We are interested in prior work demonstrating:
This is a remote contract engagement. Scope, deliverables, fees and an end date are agreed before each engagement begins. A Fleet project contact handles scoping questions and reviews submitted deliverables against the agreed requirements.
Initial hourly fees are $60–$100 USD. Rates may increase up to $150 USD based on the quality of completed deliverables. Initial fees are agreed before the engagement begins, and any rate increases are agreed before they take effect.
After an initial conversation, you may be invited to complete a take-home mechanical engineering exercise. The exercise is designed to take no more than three hours and helps us evaluate your technical approach, engineering judgment, and written explanations. Instructions and submission requirements will be provided before you begin. Submissions are used only to evaluate suitability for the engagement.
","body_markdown":"## About Fleet AI\n\nFleet AI builds training environments where AI agents practice real tasks and their work can be evaluated. Our mechanical engineering projects use realistic engineering problems to test how well those agents reason, use tools and produce valid results.\n\n## Scope of services\n\nDevelop mechanical engineering problems and reference materials for AI training and evaluation. Depending on the project, the work may cover CAD, structural or thermal analysis, fluid flow, manufacturing drawings, or design validation.\n\n## Deliverables\n\nEach project will specify the outputs to be delivered. These may include:\n\n- An engineering problem with the inputs, constraints and files needed to solve it.\n- CAD models, drawings, calculations or simulation results that establish a reference solution.\n- Written explanations of assumptions, boundary conditions, tradeoffs and limitations.\n- Criteria for evaluating a solution’s physical validity, safety and feasibility.\n- Reviews of AI-generated engineering work that identify errors and explain their consequences.\n\n## Relevant experience\n\nWe are interested in prior work demonstrating:\n\n- Professional mechanical engineering experience, including designs taken through analysis, validation or production.\n- Experience with relevant CAD or simulation tools, such as SolidWorks, Fusion, FEA or CFD software.\n- Sound judgment about physical constraints, failure modes and the limits of a model or analysis.\n- Clear technical writing that explains both the result and the reasoning behind it.\n\n## Engagement structure\n\nThis is a remote contract engagement. Scope, deliverables, fees and an end date are agreed before each engagement begins. A Fleet project contact handles scoping questions and reviews submitted deliverables against the agreed requirements.\n\n## Fees\n\nInitial hourly fees are $60–$100 USD. Rates may increase up to $150 USD based on the quality of completed deliverables. Initial fees are agreed before the engagement begins, and any rate increases are agreed before they take effect.\n\n## Express Interest\n\nAfter an initial conversation, you may be invited to complete a take-home mechanical engineering exercise. The exercise is designed to take no more than three hours and helps us evaluate your technical approach, engineering judgment, and written explanations. Instructions and submission requirements will be provided before you begin. Submissions are used only to evaluate suitability for the engagement.\n","is_active":true,"display_order":0,"created_at":"2026-02-19T06:24:11.37794+00:00","updated_at":"2026-10-14T18:02:49.908757+00:00","screening_questions":[{"id":"a1b2c3d4-e5f6-7890-abcd-ef1234567890","text":"Please provide a link to your portfolio.","type":"text","options":null,"required":false},{"id":"b2c3d4e5-f6a7-8901-bcde-f12345678901","text":"Current employer","type":"text","options":null,"required":false},{"id":"c3d4e5f6-a7b8-9012-cdef-123456789012","text":"Country of residence","type":"text","options":null,"required":true}]},{"id":"a302f51e-7393-4911-92a4-d49fc98b1d67","slug":"security-engineer","title":"Security Engineer","subtitle":"Member of Technical Staff","locations":["San Francisco, CA"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$400k","equity_range":null,"referral_bonus":null,"body":"Fleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
Fleet moves quickly because our team can trust the systems behind its work. As one of Fleet’s Security Engineers, you will own the security loop that makes that possible: seeing threats early, responding decisively, and turning every finding into a stronger system.
This is a hands-on, high-ownership role across Fleet’s cloud infrastructure and corporate systems. You’ll investigate alerts and security reports, lead the response when something matters, ship the changes that close the issue, and build automation that expands what a small security team can accomplish.
See what matters. Build and tune high-signal detections across cloud infrastructure, identity, endpoints, and business systems.
Lead the response. Investigate suspicious activity, contain threats, find root causes, and make Fleet stronger after every incident.
Turn findings into fixes. Trace security reports and vulnerabilities through the systems they affect, ship durable remediations, and verify that they work.
Harden the foundation. Strengthen cloud access, workloads, networking, secrets, telemetry, and deployment paths while preserving the speed of the teams that depend on them.
Automate the function. Use agents, APIs, code, and workflows to accelerate detection, investigation, remediation, and verification.
Experience owning detection and response, cloud security, infrastructure security, or security engineering in production.
The instincts of both an investigator and an engineer: you can follow evidence across logs, code, and infrastructure then turn what you learn into a lasting technical improvement.
Strong understanding of cloud infrastructure, identity and access, networking, logging, containers, and modern application environments.
Comfortable writing code, building integrations, and making changes through version-controlled infrastructure and configuration.
Experience using AI tools and agents to build security capabilities that are fast, reliable, observable, and auditable.
Strong judgment about blast radius, testing, rollout, monitoring, and rollback. You know how to raise the security bar while keeping the business moving.
Full-time, on-site in San Francisco, CA. Compensation of $300K - $400K salary + meaningful equity.
Salary range: $200,000–$400,000 USD annually, plus meaningful equity.
","body_markdown":"## **About Fleet**\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## **About the Role**\n\nFleet moves quickly because our team can trust the systems behind its work. As one of Fleet’s Security Engineers, you will own the security loop that makes that possible: seeing threats early, responding decisively, and turning every finding into a stronger system.\n\nThis is a hands-on, high-ownership role across Fleet’s cloud infrastructure and corporate systems. You’ll investigate alerts and security reports, lead the response when something matters, ship the changes that close the issue, and build automation that expands what a small security team can accomplish.\n\n## **What You’ll Do**\n\n- **See what matters.** Build and tune high-signal detections across cloud infrastructure, identity, endpoints, and business systems.\n- **Lead the response.** Investigate suspicious activity, contain threats, find root causes, and make Fleet stronger after every incident.\n- **Turn findings into fixes.** Trace security reports and vulnerabilities through the systems they affect, ship durable remediations, and verify that they work.\n- **Harden the foundation.** Strengthen cloud access, workloads, networking, secrets, telemetry, and deployment paths while preserving the speed of the teams that depend on them.\n- **Automate the function.** Use agents, APIs, code, and workflows to accelerate detection, investigation, remediation, and verification.\n\n## **What We’re Looking For**\n\n- Experience owning detection and response, cloud security, infrastructure security, or security engineering in production.\n- The instincts of both an investigator and an engineer: you can follow evidence across logs, code, and infrastructure then turn what you learn into a lasting technical improvement.\n- Strong understanding of cloud infrastructure, identity and access, networking, logging, containers, and modern application environments.\n- Comfortable writing code, building integrations, and making changes through version-controlled infrastructure and configuration.\n- Experience using AI tools and agents to build security capabilities that are fast, reliable, observable, and auditable.\n- Strong judgment about blast radius, testing, rollout, monitoring, and rollback. You know how to raise the security bar while keeping the business moving.\n\n## Compensation\n\nFull-time, on-site in San Francisco, CA.\n\nSalary range: $200,000–$400,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-09-08T01:22:35.562353+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"c42cec1e-dd67-41fd-8bbc-6d8f63921077","slug":"mts-synthetic-data","title":"Synthetic Data & Evals","subtitle":"Research Engineer","locations":["San Francisco","New York"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Fleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures — the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities — building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
We’re looking for a Research Engineer to build reusable and reliable systems for synthetic data and evaluations.
This work spans synthetic data and task generation, verifiers, quality measurement, and the systems that connect and orchestrate the entire loop. You will develop new methods, test them on open-ended research problems, and turn the ones that work into systems that you and other researchers can use and extend.
You will work closely with internal research teams and frontier labs, study model behavior and traces directly, and build feedback loops between research findings and system improvements.
Success means making high-quality research faster, more reproducible, and easier to scale across researchers and domains, with the longer-term goal of systems that use evaluation feedback to recursively improve their own outputs.
Build reusable and reliable systems for producing high-quality synthetic data, tasks, and verifiers across software engineering, cyber, legal, finance, healthcare, and other technical and knowledge work domains.
Leverage research signals and expert knowledge to define and enforce quality standards for each eval set—understand what “good” looks like for each domain and iterate until it is right.
Build agentic hill-climbing and feedback loops that continuously improve the systems and their outputs against those standards.
Collaborate closely with environment and research teams, frontier labs, and Fleet’s expert network to translate research needs and model failures into shared system capabilities.
Treat the systems and their outputs as products: own quality and usability end to end, including researcher education, adoption, and continuous iteration based on how the systems are used.
Strong engineering fundamentals and experience building reliable systems that others depend on. Proficient in Python or another modern backend language; comfortable across containers, distributed systems, and GPUs when the work requires it.
Strong ML fundamentals. You understand how training objectives and optimization shape model behavior, and can reason from first principles about why an approach works or fails.
Ability to turn a research prototype into a maintainable system with clear interfaces, reliable execution, and useful debugging tools.
Strong judgment about data and system quality: able to use expert knowledge and research signals to define what “good” looks like, translate ambiguous quality goals into useful signals, and diagnose why a system or its outputs are failing.
Ability to work effectively with internal research teams, domain experts, and frontier lab partners.
High ownership, low ego, and a preference for systems that are simple and boring.
We care more about demonstrated ability than credentials. Strong candidates may come from research engineering, backend or data systems, ML platforms, developer tooling, synthetic data, or applied AI.
Post-training experience and research taste. You’ve owned experiments to improve model capabilities, designed ablations to understand what worked, and can identify promising research directions while distinguishing meaningful progress from misleading results.
Hands-on experience building synthetic or domain-specific datasets and eval sets, reward models, agentic workflows, or automated quality systems.
Small, technical team.
Speed with rigor.
Truth over comfort. Honest with partners, with each other, and with ourselves.
On-site in SF or NYC.
On-site in San Francisco or New York. Highly competitive salary + equity.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures — the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities — building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## About the Role\n\nWe’re looking for a Research Engineer to build reusable and reliable systems for synthetic data and evaluations.\n\nThis work spans synthetic data and task generation, verifiers, quality measurement, and the systems that connect and orchestrate the entire loop. You will develop new methods, test them on open-ended research problems, and turn the ones that work into systems that you and other researchers can use and extend.\n\nYou will work closely with internal research teams and frontier labs, study model behavior and traces directly, and build feedback loops between research findings and system improvements.\n\nSuccess means making high-quality research faster, more reproducible, and easier to scale across researchers and domains, with the longer-term goal of systems that use evaluation feedback to recursively improve their own outputs.\n\n## What You'll Work On\n\n- Build reusable and reliable systems for producing high-quality synthetic data, tasks, and verifiers across software engineering, cyber, legal, finance, healthcare, and other technical and knowledge work domains.\n- Leverage research signals and expert knowledge to define and enforce quality standards for each eval set—understand what “good” looks like for each domain and iterate until it is right.\n- Build agentic hill-climbing and feedback loops that continuously improve the systems and their outputs against those standards.\n- Collaborate closely with environment and research teams, frontier labs, and Fleet’s expert network to translate research needs and model failures into shared system capabilities.\n- Treat the systems and their outputs as products: own quality and usability end to end, including researcher education, adoption, and continuous iteration based on how the systems are used.\n\n## What We're Looking For\n\n- Strong engineering fundamentals and experience building reliable systems that others depend on. Proficient in Python or another modern backend language; comfortable across containers, distributed systems, and GPUs when the work requires it.\n- Strong ML fundamentals. You understand how training objectives and optimization shape model behavior, and can reason from first principles about why an approach works or fails.\n- Ability to turn a research prototype into a maintainable system with clear interfaces, reliable execution, and useful debugging tools.\n- Strong judgment about data and system quality: able to use expert knowledge and research signals to define what “good” looks like, translate ambiguous quality goals into useful signals, and diagnose why a system or its outputs are failing.\n- Ability to work effectively with internal research teams, domain experts, and frontier lab partners.\n- High ownership, low ego, and a preference for systems that are simple and boring.\n\nWe care more about demonstrated ability than credentials. Strong candidates may come from research engineering, backend or data systems, ML platforms, developer tooling, synthetic data, or applied AI.\n\n## Especially Exciting\n\n- Post-training experience and research taste. You’ve owned experiments to improve model capabilities, designed ablations to understand what worked, and can identify promising research directions while distinguishing meaningful progress from misleading results.\n- Hands-on experience building synthetic or domain-specific datasets and eval sets, reward models, agentic workflows, or automated quality systems.\n\n## How We Work\n\n- Small, technical team.\n- Speed with rigor.\n- Truth over comfort. Honest with partners, with each other, and with ourselves.\n- On-site in SF or NYC.\n\n### Location\n\nOn-site in San Francisco or New York.\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-06-08T22:31:45.159664+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"66fa4764-3868-423b-bccb-0225fdb69669","slug":"software-engineer","title":"Environment Generation","subtitle":"Member of Technical Staff","locations":["San Francisco","New York"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":500,"body":"Fleet AI is an applied artificial intelligence company focused on human-AI collaboration at scale.
Our goal is to accelerate the transition to the allocation economy, a future where humans direct work rather than perform it. We build training gyms for AI agents: high-fidelity simulation environments where agents practice real tasks while humans supervise, evaluate, and steer their behavior.
As an Environment Engineer, you own environments end-to-end, from designing the env through to the first graded agent rollout. Each environment is a full-stack rebuild of how a real business actually runs, faithful to the original right down to its conventions, quirks, and how things break in production. The bar is high: an agent that masters our env should be able to do the real job. In practice, that means full-stack rebuilds of real products, instrumented with agent-legible tool surfaces, data, tasks, and programmatic verifiers that turn every rollout into reward.
You define the env, build it, design the tool contracts, generate the data, integrate expert feedback, and take the env from \"built\" to \"producing training signal.\" Your taste at every stage shapes what agents can learn.
Own a vertical end-to-end:
Build the env: research the real product and ship the full-stack app, including the data model, UI, business logic, tool surfaces legible to agents, and seed pipelines holding realistic, resettable state across millions of rollouts. The first graded rollout proves the whole stack works.
Own the hardest envs on the roadmap: multi-step tasks that need coherent state across many actions (like closing the books at quarter-end, or running a multi-week candidate sourcing pipeline), interactions with simulated stakeholders who push back (a customer disputing a charge, a hiring manager rejecting a candidate), real codebases with conventions and tech debt rather than toy repos, and open-ended design problems with no single right answer. Once these are built, much of the work is debugging: when an agent fails, distinguishing genuine environment issues from artifacts and redesigning to target deeper failure modes.
Review at scale: as Fleet's env library grows, quality only holds if review keeps pace. You'll programmatically review envs, data, and tasks for correctness and robustness, advise on data, task, and grader design, and build and maintain the review standards we need to keep scaling.
Expert feedback: work with domain experts to build, review, and QA envs, data, and tasks at volume, and ship the systems that keep that feedback loop tight.
Leverage: direct fleets of coding agents to parallelize development and raise the ceiling on what one engineer ships.
Platform: shape the shared abstractions and quality bars every Fleet environment inherits.
Are a high-agency, deeply technical full-stack builder with strong TypeScript and web fundamentals
Have product taste. You can study a real app and know which details matter to faithfully recreate it
Are comfortable working directly with domain experts (accountants, recruiters, IT admins) and turning their feedback into code
Ship fast and thrive in ambiguity. You can figure out the right approach without a detailed spec
Use AI tools daily and can effectively direct multiple coding agents in parallel
Sweat the details that make a simulation indistinguishable from the real thing
Think in terms of determinism, invariants, and contracts, not just features
Have a deep appreciation for AI research and low-level infrastructure
Thrive on routine, consistency, and long planning cycles. We ship in days, not quarters.
Want to go deep on one narrow technical problem; this role spans the stack and crosses into research, infra, and product
Need a clean handoff between spec, design, and engineering. Here, those are the same person.
Aren't excited to spend serious time communicating clearly with AI agents
Move fast without caring about what you ship. We don't trade quality for speed.
TypeScript, Svelte/SvelteKit, React, Node.js, Express, oRPC, Tailwind, SQLite/Postgres with Drizzle, Python, Docker, AWS (ECR/ECS), MCP (Model Context Protocol), Vitest, Playwright, Claude Code and multi-agent workflows.
Full-time, on-site in SF or NYC. Compensation (salary + meaningful equity) will be extremely competitive.
We believe the most important advances in AI will come from close, thoughtful collaboration between humans and machines. We own our outcomes, move with relentless speed when we build conviction, and choose truth over comfort. We don't cut corners. We cut scope.
We encourage you to apply even if you don't meet every qualification listed. We care far more about how you build and think than about credentials alone.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet AI\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. \n\nWe work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## About the Role\n\nWe are looking for engineers to build code generation pipelines that produce high quality, high fidelity RL environments at scale.\n\n\nWe come at this from two directions:\n\n- **Terraforming the internet**:\nIn the same way Google indexed the web, we crawl it ourselves, looking for environments that teach agents useful behavior. Every part of the digital world, from enterprise apps to YouTube videos to Kaggle competitions to prediction markets, can become a pipeline that builds rich simulations for agent training.\n- **Learning from agent experience**. While we use up the internet, a new data source is growing exponentially. Agents are working in the real world and producing huge volumes of experience. We want to learn from it with open-loop simulation (as is common in self-driving), synthesize environments from it, and turn the real world into a verification signal.\n\nWe obsess over quality and realism. A bug in an environment teaches the agent the wrong behavior, and a bug in our generator repeats it across many environments.\n\nThere are a lot of unknowns, so you will be expected to propose and evaluate different approaches.\n\n\n## What You'll Work On\n\n- Building and scaling agentic systems that generate large, complex software with minimal human input\n- Making those systems reliable over long running work, and debugging them when they aren't\n- Holding a very high bar for quality and correctness in everything they produce\n- Prototyping and evaluating new approaches to problems nobody has solved yet\n\n\n## What We're Looking For\n\n- Prior experience with code generation and agent harnesses\n- You've worked on coding agents that do long, complex work reliably\n- You understand coding agent failure modes very well\n- You are comfortable working on nebulous problems\n- You care a lot about quality and know how your work can break\n- Bonus:\n - Experience across the end to end SDLC\n - You've built automated systems that judge the quality of complex software and exhaustively find bugs\n\n## How We Work\n- Small, technical team.\n- Quick proofs of concept to find good approaches.\n- Speed with rigor.\n- Truth over comfort. Honest with partners, with each other, and with ourselves.\n- On-site in SF or NYC.\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":0,"created_at":"2026-03-13T21:05:58.581838+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"00c68b47-1bf5-4c7b-86bc-76e3a7873046","slug":"research-scientist-post-training","title":"Post-Training","subtitle":"Research Scientist","locations":["San Francisco"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Fleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
\nWe're looking for a Research Scientist to work on agentic RL post-training: the recipes that turn environment interaction into model capability. Tasks and environments are half the problem; the other half is training on them well. You'll train specialized agent models against Fleet's environments, scale the RL runs behind them, and own the algorithmic work that makes those runs pay off.
\nSan Francisco (on-site). Highly competitive salary + equity.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"### About Fleet\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n### About the Role\n\nWe're looking for a Research Scientist to work on agentic RL post-training: the recipes that turn environment interaction into model capability. Tasks and environments are half the problem; the other half is training on them well. You'll train specialized agent models against Fleet's environments, scale the RL runs behind them, and own the algorithmic work that makes those runs pay off.\n\n### What You'll Work On\n\n- RL post-training recipes for multi-turn agents across modalities (tool use, browser use, terminal use)\n- Denser reward signals: verifiers, process supervision, richer feedback than binary outcomes\n- Scaling RL runs: larger models, longer horizons, many environments in a single run\n- Training specialized models that demonstrate what Fleet's environments teach\n- Algorithmic innovations in agentic RL, published where it makes sense\n\n### How We Work\n\n- You carry your own research threads that lead to concrete artifacts: a paper, benchmark release, training recipe, or open-source code\n- Direct collaboration with frontier labs\n\n### What We're Looking For\n\n- Hands-on experience post-training LLMs with RL, end-to-end\n- Publication record at top venues (NeurIPS, ICML, ICLR, or equivalent), or shipped training runs at comparable scale\n- High ownership: you manage your own research threads and are measured by what you produce\n\n### Location\n\nSan Francisco (on-site).\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":2,"created_at":"2026-07-23T14:35:18+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"50a82de0-a240-4b6d-a936-70acd86848c4","slug":"research-scientist-oversight","title":"Oversight","subtitle":"Research Scientist","locations":["San Francisco"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Fleet studies how environments produce intelligence: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We build the benchmarks that expose where frontier models break and the training that closes the gap. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
\nWe're looking for a Research Scientist to work on scalable oversight for agents: detecting misbehavior like reward hacking and deception, understanding why it happens, and keeping verification reliable as models outgrow the humans and automated systems checking them. Because Fleet builds sandboxed environments and training tasks for agents, our task factory will be your controlled laboratory where you can intervene, attribute behavior to its cause, and feed what you learn back into improvements. You'll own research threads that ship as papers, benchmarks, training gyms, models optimized for efficient oversight, or production systems, in direct collaboration with frontier labs.
\nSan Francisco (on-site). Highly competitive salary + equity.
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"### About Fleet\n\nFleet studies how environments produce intelligence: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We build the benchmarks that expose where frontier models break and the training that closes the gap. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n### About the Role\n\nWe're looking for a Research Scientist to work on scalable oversight for agents: detecting misbehavior like reward hacking and deception, understanding why it happens, and keeping verification reliable as models outgrow the humans and automated systems checking them. Because Fleet builds sandboxed environments and training tasks for agents, our task factory will be your controlled laboratory where you can intervene, attribute behavior to its cause, and feed what you learn back into improvements. You'll own research threads that ship as papers, benchmarks, training gyms, models optimized for efficient oversight, or production systems, in direct collaboration with frontier labs.\n\n### What You'll Work On\n\n- Scalable oversight: methods that keep verification reliable as tasks get longer and models get stronger.\n- Monitoring: detecting reward hacking or undesirable behavior patterns in our agent traces.\n- Failure mode analysis and attribution: tracing agents' mistakes and misbehaviors upstream to root causes in data, environments, or model weights.\n- Turning oversight into training signal: better verifiers, process supervision, and models trained to catch what humans miss.\n\n### What We're Looking For\n\n- Publication record at top venues (NeurIPS, ICML, ICLR, or equivalent), or equivalent shipped research, in oversight, monitoring, evaluations, alignment, interpretability, AI control, or RL.\n- Strong empirical instincts; comfortable with LLM APIs, harnesses, and large trace datasets.\n- High ownership: you run your own research threads and are measured by what you ship.\n\n### Location\n\nSan Francisco (on-site).\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":3,"created_at":"2026-07-24T08:46:10.356114+00:00","updated_at":"2026-10-13T20:54:29.658222+00:00","screening_questions":[]},{"id":"f2fcc886-8bb4-48fa-93f7-eb33974fd456","slug":"head-of-talent","title":"Founding Recruiter","subtitle":"","locations":["San Francisco"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Fleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.
Fleet's Founding Recruiter will build the recruiting function for our next stage of growth. You'll run searches end to end while building the systems and interviewing culture that can scale with Fleet.
Partner with leadership to turn company priorities into the hiring plan and recruiting flows required to grow Fleet from roughly 50 to 80+ over the next few months
Own recruiting end to end, from role kickoff and sourcing through interviews, closing, and onboarding
Keep the ATS accurate, identify where the hiring process is slowing down, and bring recommendations to leadership
Increase hiring throughput while decreasing the engineering hours required per hire, without lowering the bar
Build the recruiting function and determine the right mix of internal capacity, tooling, and external agencies
Flat, high-ownership team
Speed with rigor
Truth over comfort
Five days per week in office in San Francisco
Experience auditing and owning the hiring process, from company-wide planning to granular operations
Strong candidate intuition across technical and nontechnical roles
A builder's willingness to source, run introductory calls, and manage the ATS before hiring a team around you
Willing to debate hiring philosophy and interview design, then adjust as the team's shared goals evolve
Experience managing candidate relationships from initial outreach through their first quarter at the company and beyond
Experience as the first or an early recruiter at a fast-growing technology company
Experience selecting, managing, and evaluating external recruiting agencies
Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet\n\nFleet studies how environments produce intelligence. We believe intelligence is an emergent property of environmental pressures: the environment determines what capabilities develop, what behaviors survive, and what \"good\" looks like. We work with frontier labs on post-training across modalities, building benchmarks that expose where frontier models break, training recipes that close those gaps, and scalable oversight for long-horizon agents. Backed by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## About the Role\n\nFleet's Founding Recruiter will build the recruiting function for our next stage of growth. You'll run searches end to end while building the systems and interviewing culture that can scale with Fleet.\n\n## What You'll Do\n\n- Partner with leadership to turn company priorities into the hiring plan and recruiting flows required to grow Fleet from roughly 50 to 80+ over the next few months\n- Own recruiting end to end, from role kickoff and sourcing through interviews, closing, and onboarding\n- Keep the ATS accurate, identify where the hiring process is slowing down, and bring recommendations to leadership\n- Increase hiring throughput while decreasing the engineering hours required per hire, without lowering the bar\n- Build the recruiting function and determine the right mix of internal capacity, tooling, and external agencies\n\n## How We Work\n\n- Flat, high-ownership team\n- Speed with rigor\n- Truth over comfort\n- Five days per week in office in San Francisco\n\n## What We're Looking For\n\n- Experience auditing and owning the hiring process, from company-wide planning to granular operations\n- Strong candidate intuition across technical and nontechnical roles\n- A builder's willingness to source, run introductory calls, and manage the ATS before hiring a team around you\n- Willing to debate hiring philosophy and interview design, then adjust as the team's shared goals evolve\n\n## Nice to Have\n\n- Experience managing candidate relationships from initial outreach through their first quarter at the company and beyond\n- Experience as the first or an early recruiter at a fast-growing technology company\n- Experience selecting, managing, and evaluating external recruiting agencies\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":4,"created_at":"2026-08-07T23:19:06.836817+00:00","updated_at":"2026-10-14T19:00:37.556756+00:00","screening_questions":[]},{"id":"2eb60d17-2043-48ee-9839-5881d17e2ce6","slug":"research-scientist-safety-alignment","title":"Safety and Alignment","subtitle":"Applied Research Scientist","locations":["San Francisco"],"work_type":"On-site","employment_type":"Full-time","pipeline_type":"fulltime","salary_range":"$200k–$600k","equity_range":null,"referral_bonus":null,"body":"Salary range: $200,000–$600,000 USD annually, plus meaningful equity.
","body_markdown":"## About Fleet\n\nFleet studies how environments produce safe intelligence. We believe safety and economic value increasingly reinforce each other: as agents take on more consequential work, their value depends on whether they can be trusted with that responsibility.\n\nAs reinforcement learning takes a larger role in post-training, the design of tasks, environments, and feedback becomes increasingly consequential for downstream alignment. RL can reinforce sound judgment, but it can also amplify unintended behaviors. We believe studying and scaling the data and training signals that produce more aligned agents is one of the most impactful ways to contribute to AI safety.\n\nBacked by Sequoia Capital, Menlo Ventures, BCV, and SV Angel.\n\n## About the Role\n\nWe’re looking for an Applied Research Scientist to work on robust reward design, implement comprehensive telemetry, and design production systems to help ensure Fleet's RL tasks make capable agents more aligned. You’ll work with data and model trajectories to help develop oversight systems to detect misaligned behaviors. You’ll also work directly with safety researchers at frontier labs to help co-design new RL environments and tasks for downstream alignment training and evaluation.\n\n## What You’ll Work On\n\n- **Reward Design and Telemetry at Production Scale.** Study how task and verifier design affect model-alignment incentives, then improve both.\n - **Off-target behaviors from task assumptions and incentives.** Investigate how unstated behavioral constraints, unsolvable tasks, and rewards that ignore how success is achieved affect how models behave during evaluation, across many different domains.\n - **Telemetry and Verifier Generation Pipelines.** Use the insights learned from the above to build 1) the telemetry needed for customers to have complete observability over undesirable model behaviors and 2) the scalable pipelines needed for producing verifiers that translate recorded measurements into qualitative behavioral categories.\n - **Production Oversight Systems.** Build out rigorous trace-based analysis systems that measure agent misbehavior across all internal Fleet evaluations.\n- **Safety evaluations and benchmark analysis.** Develop rigorous safety measurements and assess the risks in existing Fleet tasks and external benchmarks.\n - **Safety benchmarks.** Build held-out and adversarial evaluations to determine the safety profile of existing tasks. Audit Fleet tasks, data, and graders for contamination, poisoning, ambiguity, and shortcuts.\n - **External benchmark analysis.** Examine widely used benchmarks, datasets, and task designs for incentives that could reinforce unsafe behavior.\n- **Specialization in high-impact domains.** Get involved in domain-specific safety research with frontier labs, including cybersecurity. Work with experts to identify consequential failure modes and design curricula, tasks, and evaluations aimed at preventing incidents like the Hugging Face breach.\n\n## What We’re Looking For\n\n- A willingness to work directly with the data: inspect examples and trajectories, question labels and graders, investigate anomalies, and trace a surprising result to its cause.\n- Strong practical engineering skills—you should be able to build the systems needed to test and scale your ideas.\n- A record of strong empirical experience in alignment, model behavior, model evals, or a closely related field.\n\nPrior research experience via safety fellowships (e.g. MATS, Astra), a safety organization (e.g. METR, Redwood, Epoch AI), or academia is helpful but not necessary.\n\n## Location\n\nSan Francisco (on-site).\n\n## Compensation\n\nSalary range: $200,000–$600,000 USD annually, plus meaningful equity.","is_active":true,"display_order":5,"created_at":"2026-09-13T21:31:28.928132+00:00","updated_at":"2026-10-14T19:00:15.938226+00:00","screening_questions":[]},{"id":"51c3befe-b01e-4d6d-aa1c-abeaee503264","slug":"domain-expert-legal","title":"Legal Specialist — Contract Engagement","subtitle":null,"locations":["Remote (US)"],"work_type":"Remote","employment_type":"Contract","pipeline_type":"contractor","salary_range":"$150–$200/hr","equity_range":null,"referral_bonus":500,"body":null,"body_markdown":"## About Fleet AI\nFleet AI builds training environments where AI agents practice real tasks and their work can be evaluated. Our legal projects use realistic matters and legal work products to evaluate how well those agents research, reason, and support their conclusions.\n\n## Scope of services\nDevelop legal tasks and reference materials for AI training and evaluation. Each engagement will specify the practice area, jurisdiction, outputs, and acceptance criteria.\n\n## Deliverables\nEach project will specify the outputs to be delivered. These may include:\n- A legal assignment with a case file and supporting documents.\n- A reference memo, brief, diligence report, or other work product that answers the assignment.\n- Evaluation criteria grounded in the case file and relevant legal authorities.\n- Reviews of AI-generated work identifying errors in legal reasoning, factual support, citations, or conclusions.\n\n## Relevant experience\nWe are interested in prior work demonstrating:\n- Experience in US legal practice, ideally as a senior paralegal or attorney, with substantive research, drafting, or document review experience. Active bar admission is a plus but not required.\n- Ability to explain why a conclusion is supported, where the evidence is incomplete, and why an error matters.\n- Clear written analysis grounded in the relevant facts and jurisdiction.\n\n## Engagement structure\n- **Format:** Remote, part-time contract.\n- **Initial term:** The first phase runs through October 31, 2026, as part of a multi-phase, long-term program.\n- **Location:** United States only. Fleet is unable to sponsor work authorization in the United States or elsewhere.\n- **Terms:** Scope, deliverables, fees, and an end date are agreed before work begins, and any extension is agreed in writing. A Fleet project contact handles scoping questions and reviews submitted deliverables.\n\n## Fees\n- **Rate:** $150–$200 USD per hour.\n- **Increases and bonuses:** Top performers may earn higher rates based on quality, consistency, and throughput. Per-task bonuses may be available for especially challenging, high-quality work.\n- **Tools:** Reasonable AI tool subscriptions are reimbursed.\n\n## Express Interest\nApply using the form on this page. The selection process is a short, unpaid online assessment in which you review a legal task. There is no interview. Instructions and expected completion time are provided before you start.\n\n**Referrals:** $500 once someone you refer completes 40 hours of work.","is_active":true,"display_order":6,"created_at":"2026-09-25T10:55:08.381719+00:00","updated_at":"2026-10-21T21:25:02.926291+00:00","screening_questions":[{"id":"4b6def57-c547-43d8-a1f2-c70dc47f997d","text":"Current employer","type":"text","options":null,"required":false},{"id":"72735e85-4d9f-482d-95ae-9377517031f8","text":"Country of residence","type":"text","options":null,"required":true},{"id":"2e458294-f4cc-4906-923e-412b78609fc4","text":"Are you licensed to practice law?","type":"select","options":["Yes, in the US","Licensed outside the US only","No"],"required":true},{"id":"0460cdb1-1494-4620-9456-841621b1ecfa","text":"Which best describes your current legal role?","type":"select","options":["Practicing attorney","In-house counsel","Paralegal / legal assistant","Legal operations / contracts / compliance","Law student or recent JD (not yet admitted)","Other"],"required":true},{"id":"f42d5a7c-3ac7-40c1-b8d0-94a2b9331acb","text":"Years of post-JD legal experience","type":"select","options":["0-2","3-5","6-10","10+"],"required":true},{"id":"9ad496ee-038f-432c-a510-631087f0d971","text":"Primary practice area","type":"select","options":["Litigation","Corporate / M&A","Intellectual property","Employment","Regulatory / compliance","Other"],"required":true},{"id":"df9fd005-6d54-49ad-b828-3814b7149b63","text":"Hours per week you can commit through Oct 31","type":"select","options":["0","1-20","21-40","41-60","61-80","80+"],"required":true},{"id":"6b4f787f-50c3-4bf7-9fce-b898d8cfe7e7","text":"Hours per week you expect to be available Nov 1 - Dec 31 (Q4)","type":"select","options":["0","1-20","21-40","41-60","61-80","80+"],"required":true},{"id":"cda8ca3a-2a81-4202-8578-26cfa09362a8","text":"Hours per week you expect to be available in early Q1 2027","type":"select","options":["0","1-20","21-40","41-60","61-80","80+"],"required":true}]}]