Built a RAG-based
knowledge assistant that connects to private documents, retrieves relevant
context, and generates grounded answers for users. The system includes document
ingestion, chunking, embeddings, vector search, and LLM-based response
generation
