I design and build Retrieval-Augmented Generation (RAG) systems that let your users ask natural-language questions over your own documents, knowledge base, or product data — and get accurate, source-grounded answers. This is the AI feature that actually saves teams hours of search every day. What's included: — Document ingestion pipeline (PDF, CSV, Notion, web pages, etc.) — Embedding generation + vector store setup (Pinecone, Weaviate, or pgvector) — LlamaIndex or LangChain retrieval chain with reranking — Clean API endpoint to query your knowledge base — Hybrid search (semantic + keyword) for better accuracy — Source citation in responses — Optional chat UI (if needed)