Work

Four cases. Each one answers the same four questions: what was the problem, what did I do, what did I measure and what did I learn.

1. RAG in productionDocument search with embeddings and a reranker, and an evaluation with fixed scores instead of testing by feel.MRR 0.950on a fixed set of evaluation questions
2. Working with AI agents in a teamShared rules, a hook that blocks dangerous git commands, a model per task and a fixed verification step.
3. Self-hosting on a NASMy own apps live with Docker, Forgejo and a reverse proxy, plus a security mistake I found and closed myself.
4. Laravel APIs with Nuxt frontendsThe daily work, an API in Laravel with a frontend in Nuxt, described in general terms and without client names.