Technical insights from real KYFEX projects
How we built a production RAG pipeline using managed AI services and vector search. Architecture decisions, lessons learned, and performance benchmarks.