Staff Research Engineer, Multi-Agent Scaling

Highlights

• Study how large agent teams scale across size, compute, and task horizon.
• Build the platform Anthropic uses to run and measure very large agent teams.
• Design long-horizon evals that distinguish real teamwork gains from artifacts.

What they offer

• Competitive compensation and benefits
• Optional equity donation matching
• Generous vacation and parental leave
• Flexible working hours
• Lovely office space to collaborate with colleagues

Ideal candidate

A generalist research engineer who loves both rigorous experimentation and building reliable systems, thrives on vague questions, thinks quantitatively, and cares about AI safety and societal impact.

Requirements

• Significant software engineering, ML or research engineering experience
• Owned something substantial end-to-end, such as a large system, evaluation, agent product, or research project
• Genuinely enjoys both research and engineering work
• Thinks quantitatively about complex systems and questions data
• Can work from a vague question rather than a spec
• Results-oriented with bias towards flexibility and impact
• Clear written and verbal communication
• Cares about the societal impacts of work