Production RAG + Azure OpenAI platform
The flagship: a live retrieval system for one client across two contracts and a year of continuous work. Rebuilt ingestion around structure-aware chunking, moved retrieval to hybrid vector + BM25 with reranking, added a citation contract — then hardened the live system against prompt injection and data leakage once real users arrived.
- 1,970hours, both phases
- $80.6Kbilled to date
- 4.9★phase-one rating