In short
ML engineer at a well-funded stealth frontier AI lab in San Francisco designing and shipping agentic systems on top of frontier LLMs. Day to day: build agent architectures (planning, tool use, memory, verification), create evaluation loops, and own the data/knowledge layer agents depend on.
About the role
Role: Member of Technical Staff - Machine Learning Engineer
Location: San Francisco
We are working with a well funded frontier AI lab, currently operating in stealth, who are building agentic systems around frontier LLMs. This role blends deep ML intuition with strong systems engineering. You will join a small but elite and fast growing technical team, which offers real ownership of end-to-end outcomes and the chance to join a lab which is building the next generation of AI systems.
The role
- Design, build and iterate on agent architectures: planning, tool use, routing, memory, verification, and long-horizon execution
- Create evaluation loops for agent reliability
- Own the data/knowledge layer that agents depend on
- Improve performance and reliability
- Collaborate with post-training and evaluation to align model behavior with the enviroments the agents run in
What they're looking for
- Strong intuition for machine learning and LLM systems.
- Experience building and deploying LLM-powered applications or agents under real-world reliability and performance constraints.
- Ability to reason from first principles and debug complex, stochastic, and distributed system failures.
- Comfort operating across research, product, and infrastructure boundaries in fast-moving environments.
- Excellent software engineering fundamentals, including ownership of production-quality codebases and collaborative development practices.
What is on offer:
- Top-of-market compensation + meaningful equity
- Ownership over high-leverage systems at the frontier
- Strong compute/tooling support
- Benefits: medical, dental/vision, 401(k), unlimited PTO.
If you are interested please apply or reach out to me directly.
Requirements
- Strong intuition for machine learning and LLM systems
- Experience building and deploying LLM-powered applications or agents under real-world reliability and performance constraints