RAG (Retrieval-Augmented Generation)
Definition
A technique where AI searches through a database of documents to find relevant information before generating its answer. This makes responses more accurate and grounded in real data instead of relying solely on training data.
Example
A customer support chatbot using RAG searches your help docs first, then generates an answer based on what it found, rather than guessing from its general knowledge.