I build tools and benchmarks for understanding how AI agents reason. My work is focused on LLM post-training, reinforcement learning, inference-time learning, and agentic AI systems.

Find me on GitHub and LinkedIn.