Senior Research Scientist, Agent Post-Training for Computer Use
AI research and deployment company with a team post-training frontier agent models
| Employer | AI research and deployment company with a team post-training frontier agent models (not named) |
|---|---|
| Location | San Francisco, CA (Bay Area) · On-site |
| Type & level | Full-time · Senior |
| Pay | USD 380k–500k base Source: pay range stated in the source posting |
| Visa path | O-1A fit |
| Role family | Foundation Model Research Scientist |
About the role
This team trains the models behind agents that operate computers, including browsers and desktops, and carry out long tasks reliably. The work covers the data, environments, graders and training methods that shape agent behavior, then carries results into major training runs and shipped products. The role exists to turn observed model failures into training signal and measurable improvements.
What you'll do
- Design and run experiments that improve agent behavior for complex computer use across desktop and browser.
- Own improvements to the post-training stack, including RL, data pipelines, graders, reward signals and diagnostics.
- Build evals and environments that expose the next model failures and convert them into training data or research directions.
- Partner with product teams to turn user needs into model improvements.
- Work on early-training and alignment interventions such as data mixtures, objectives and synthetic data.
- Debug hard failures in near-shipped models and help decide what is ready for major training runs.
Requirements
- Strong fundamentals in machine learning, software engineering, systems or statistics.
- Hands-on experience with LLMs, RL or RLHF, post-training, evals, synthetic data or tool-using agents.
- Ability to move from a vague behavioral problem to a hypothesis, pipeline, experiment and decision.
- Comfort working across research, product, infrastructure, data and safety teams.
- Willingness to build reliable training systems and processes when the team needs them.
Nice to have
Pay & benefits
USD 380k–500k base. The posting does not list benefits.
Pay source: pay range stated in the source posting. Market data, not an offer from O1DMatch.
Visa path
People in post-training research often support an O-1A case with published research, citations, contributions to widely used models or benchmarks, peer review and high compensation. The posting states no citizenship restriction. An employer like this or a U.S. agent can file the petition, and approval depends on the individual's evidence.
How to get matched
- Take the free eligibility checkAn AI review of your background against the eight O-1 criteria, at check.o1dmatch.com.
- Create your profile with code WORKWITHUSIt unlocks the Starter plan at no cost. Tell us your field and the work you do today.
- Upload your evidenceEach document is sorted under the criterion it supports, and your criteria map shows what is still missing.
- Get matchedWe match strong profiles with U.S. employers hiring for roles like this one. Gigs you confirm with verified companies count toward your itinerary, and at a score of 75 we connect you with an immigration attorney.
This listing describes a real U.S. posting so you can see what employers ask for. O1DMatch does not represent that employer and does not hold this opening; applying matches you with employers hiring for this kind of role. No visa outcome is promised.