Fulcrum
ActiveThe agentic debugger for AI systems
About
We build red-teaming agents that help developers fix their RL environments and improve their agents. Our customers are tired of being bullshitted by their agents.
Founders · 2
MIT CS & Math ’24. MEng AI candidate under Jacob Andreas (LINGO) group. Best Paper—NeurIPS ML Safety ’22; follow-up at NeurIPS ’23 and benchmark paper accepted to NeurIPS ’24. Emergent Ventures Fellow ’25. Founder of MAIA (MIT AI Alignment). Research stints at Redwood Research and Truthful AI. USABO Nationals finalist (camper), two-time AIME qualifier.
working on better human AI collaboration - building oversight interfaces for AI agents. I did ML research on in context learning and scaling laws, studied math and CS at MIT, into writing and climbing. From Paris. Website: https://uzpg.me
Launch
Fulcrum - for those who are tired of being bullshitted by their agents
Fulcrum builds software to debug AI agents and their environments. It offers monitoring, red-teaming, and explorable reports to identify why agents fail, detect bugs or reward hacking, and diagnose issues in RL environments and agent traces.
Related startups
Red teaming for AI agents
AI coworkers for fraud and risk teams.
