Fulcrum
The agentic debugger for AI systems
What they do
Developers lack reliable ways to detect and diagnose failures, exploits, and misleading behavior in reinforcement learning environments and agents.
Fulcrum provides agentic red-teaming/debugging agents that test RL systems, surface issues, and guide fixes to improve agent performance and robustness.
AI/ML developers and teams building and deploying reinforcement learning agents and environments.
Founders
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
Company timeline
- Founded2025
- YC batchSummer 2025
- StatusActive
- Team size2 people
- LocationSan Francisco, CA, USA
Similar startups
