Applied AI Researcher, Agent Systems & Evaluation

Highlights

• Build the most rigorous closed-loop evaluation system for AI work.
• Test agent hypotheses against genuine production traffic from month one.
• Own evaluation pipelines end to end, from data to iteration signal.

What they offer

• Annual performance bonus
• Equity
• Competitive benefits package

Ideal candidate

Evidence-driven researcher who distrusts intuition and lives to prove what works in production. You want your agent evaluations running for hundreds of engineers within weeks.

Requirements

• Real engineering background with strong Python and production systems experience
• Direct experience building, evaluating, or studying LLM agent systems
• Ability to stand up infrastructure for own experiments and work with data
• Vision-language model experience is a strong plus