Currently building PDF Recorder, a native macOS app for recording narrated PDF presentations.
I also work on AI agents, evaluations, and inference systems.
Recently:
- CocoIndex: async iterator cleanup fix, merged.
- kvcached: sleep/wake concurrency fix, merged.
- LM Evaluation Harness: metric-order scoring fix (open PR) and a reproducible scoring experiment.
- SGLang: Hunyuan tool-call parsing fix (open PR) and a reproducible experiment.
- τ²-bench: checkpoint recovery patch (open PR) and fault-injection results.
More contributions and experiment notes, including OpenEvolve and Inspect AI.

