Techniques & Methods
RAG (Retrieval-Augmented Generation)
Definition
A technique that enhances AI responses by first retrieving relevant information from an external knowledge base, then using that information as context for generating answers. RAG helps reduce hallucinations and keeps responses grounded in specific data.
In Plain English
Giving the AI a reference book before asking it questions. Instead of relying on memory alone, RAG lets the AI look up real information first — like how a student performs better on an open-book exam than a closed-book one.