Software engineer working on AI evaluation and reliability. I build the harnesses and release gates that decide whether a change to an AI system is ready to ship — and check that the evaluations themselves measure something.
Three of these are on PyPI: agent-release-gates · redteam-foundry · telemeval Most of the rest run as live services — links in each repo.
Available full time from October 2026.



