
Compensation
Salary undisclosedDescription
About the Role
As the Founding ML Research Engineer on a consumer AI assistant team, you will lead hands-on research and engineering to improve how the assistant responds, uses tools, and acts safely. Your work will shape product quality while balancing reliability, latency, and model costs.
What You'll Do
Build evaluation systems that turn real conversations and user preferences into trustworthy measures for product decisions.
Post-train models using supervised fine-tuning, preference tuning, and distillation, then ship the improvements to production.
Develop and evaluate model routing that uses smaller models for routine requests and more capable models when needed.
Manage the tradeoffs among response quality, latency, and cost across the product.
Improve agent reliability on websites without APIs, including navigation, sign-in, and checkout flows.
Test for adversarial inputs, data exposure, and unintended actions, then build safeguards and evaluations.
Apply ML to memory, retrieval, tool selection, and decisions about when to act or ask the user.
Review real conversations, identify opportunities for improvement, and help set research direction as the team grows.
What We're Looking For
At least 4 years of ML or software engineering experience in an AI lab, AI product company, or equivalent environment.
Hands-on experience fine-tuning or post-training large language models and shipping results to production.
Experience designing evaluations, such as LLM judges, human preference data, or traffic replay, that support product decisions.
Background in one or more areas: red teaming, browser or computer-use agents, model routing, or production agents with tool use and long-running memory.
Strong Python skills and familiarity with the ML stack, plus comfort using TypeScript for product code.
A founding-team mindset, with the ability and interest to take ownership and set research direction.
Compensation & Benefits
Annual salary range: $200,000 to $300,000. Visa sponsorship is not available.
Location
Full-time, on-site in San Francisco, California.
Stack
- Posted
- Oct 14, 2026
- Last seen
- Oct 14, 2026
- First seen
- Oct 14, 2026


