AI Evaluation Expert and Domain Reviewer Digis, for OpenAI AI Lab
I design software-development evaluation tasks for OpenAI AI Lab, including research prompts, required file artifacts, and source-backed binary rubrics. I have delivered tasks covering Next.js and Vercel database choices, Next.js cookies and CSRF, and Core Web Vitals diagnosis.
I test models to identify genuine capability gaps, review other experts' submissions, and validate artifacts and formulas directly. I also built delivery tooling including a rubric linter, a nine-gate workflow, and Claude Code skills.