Design a high-performance RAG (Retrieval Augmented Generation) architecture for a company’s internal documentation. Suggest a tech stack (Vector DB, Embedding model, LLM) and explain how to handle document chunking and metadata filtering for better retrieval accuracy.