Retrieval-Augmented Generation (RAG) integrates a vector store for embeddings with LLM queries to enhance responses via relevant external data.
