Thoughts, experiments, and resources on AI's role in software testing.
A guide to building evaluation (eval) systems for AI and LLM powered products, written to take someone from zero prior exposure through to a working, defensible evaluation practice.
A two-phase SDET prompt for Claude that forces an explicit requirements gap analysis - functional, boundary, security, and integration - before it's allowed to generate a lean, fully-traceable 12 to 20 case test suite with strict assertion and automation-readiness rules.