AI Product Engineer (LLM Systems)

Emanuele Czofei is an AI Product Engineer focused on building production LLM systems with a strong emphasis on measured quality. Their recent work includes developing a Spanish occupational-safety RAG assistant utilizing hybrid retrieval (BM25 + voyage-3 + rerank-2.5), achieving high performance metrics (Hit@8 0.98, MRR 0.955). They have expertise in LLM evaluations, having rebuilt an LLM-as-judge layer to achieve perfect TPR/TNR on faithfulness with meta-evaluation gating, and integrating these evals into CI/CD pipelines to prevent regressions. Another key project is RepoDoctor, a LangGraph agent capable of analyzing repositories, detecting tech stacks, running tests/lints in sandboxed environments, and providing structured diagnoses. Previously, they shipped auth flows using Next.js/TypeScript. Emanuele is seeking a full-time role to ship LLM features end-to-end, covering retrieval, agents, evaluations, and the surrounding product surface. They are open to EU contracts or US contracts via EOR, and can also invoice B2B as a registered freelancer.