by Michele Laurelli
A technique that enhances LLM outputs by retrieving relevant information from external knowledge bases.
RAG combines retrieval systems with generative models. First retrieves relevant documents, then uses them as context for generation. Reduces hallucinations and enables knowledge updates without retraining.
Question answering with document retrieval
Customer support chatbots
Research assistants