#software-engineering

Articles tagged with #software-engineering

Oct 23, 2026
Oct 23, 20269 min read
Your AI Evaluation Can Pass Every Test and Still Mislead You
Your AI Evaluation Can Pass Every Test and Still Mislead You Imagine shipping an AI feature after every automated test passes. The test suite is green. The evaluation script produces a score. A pairwi
9 min read
Jun 11, 2026
Jun 11, 20266 min read6 views
I turned one LLM call into a multi-agent system — and only real testing caught the bugs that mattered
Building an AI wellness coach for exhausted parents, on a zero-dollar budget, taught me the difference between a model call and a system.
6 min read6 views