sainath User

I’m a Team Lead of an MSME-funded (9 Lakhs) live healthcare platform serving real doctors and patients, specializing in Python backend engineering, LLM pipelines, and scalable AI application development. I built and deployed production-grade AI systems, including a multi-tenant RAG SaaS with sub-400ms hybrid retrieval and clinical decision support using LLMs, with a strong focus on reliability and tenant isolation.

In my projects, I engineered production systems end-to-end: DocuMind Enterprise uses 3-stage retrieval (Hybrid Search/RRF + Cohere Reranking + SSE streaming) with strict tenant isolation via scoped Pinecone namespaces and PostgreSQL row-level filtering, plus LLM-as-a-Judge quality evaluation and a Semantic Redis Cache for faster repeated queries. I also led backend work on Doc2U (25+ RESTful API endpoints, JWT-based multi-role authentication, and AI-driven smart triage) and built a RAG-based rare disease diagnosis pipeline over 1,500 records—reaching Top-1: 93% match accuracy as a proof-of-concept.