Hi there! I work on tools to write more powerful, legible, and ergonomic software specifications, and therefore more correct software writ large.

I work at Antithesis, where I co-created Hegel, a universal property-based testing protocol to bring PBT to all developers.

I also maintain Hypothesis, Python's property-based testing library. In my time as a maintainer, I've done foundational work on Hypothesis, including building a new core representation and making Hypothesis thread-safe.

Together with Zac, I co-founded HypoFuzz, providing coverage-guided fuzzing of Hypothesis tests, autoscaling compute of those distributed fuzz workers, and visual fuzzing observability data.

As for academia: I co-authored this NeurIPS workshop paper in October 2025 with Anthropic. I released the Hypothesis Corpus dataset in April 2026, the largest ever set of software properties and their runtime information. And I was a participant in both iterations of METR's widely cited study on AI productivity.