Researcher, Agents for Automated Discovery

About the Company

We're building autonomous research agents for recursive self-improvement (multi-agent systems that propose, run, and analyze machine learning experiments). We're a small team based in San Francisco, on-site.

Responsibilities

• Design methods to improve agent planning, task decomposition, tool use, and error recovery
• Develop multi-agent coordination patterns for context sharing and labor division
• Build and maintain evaluation frameworks for open-ended agent capabilities
• Run rigorous experiments including controls, ablations, and statistical significance testing
• Co-design agent architectures with engineering and ship methods into production
• Stay current with agentic ML, planning, RL, and tool-use literature

Requirements

• 5+ years of hands-on research experience in industry or academia
• Strong track record of ML research focusing on agents, RL, LLMs, or multi-agent systems
• Experience designing and running end-to-end experiments at scale
• Proven ability to build evaluation frameworks for non-standardized capabilities
• Published research at NeurIPS, ICML, ICLR, COLM, RLC, or comparable venues
• Strong written communication skills
• Ability to thrive in ambiguous, open-ended research environments

Preferred Qualifications

• PhD in ML, statistics, CS, or adjacent fields
• Published research on agentic systems, long-horizon planning, or self-improvement
• Open-source contributions to the agentic ML ecosystem
• Experience with reasoning models, chain-of-thought, or supervised fine-tuning
• Background in evaluation methodology for unbenchmarked capabilities