What Is RAG? | Boostora Ink

RAG, or retrieval-augmented generation, lets an AI answer using retrieved documents or data.

RAG stands for retrieval-augmented generation. It is a pattern where a system first retrieves relevant information from a knowledge source, then gives that information to a model so it can answer with context.

In practice

RAG is common for support chatbots, internal assistants, documentation search, policy Q&A, and product knowledge. It helps the AI use information that was not in the model training data.

What to watch

RAG does not make answers automatically correct. Retrieval can find the wrong chunks, miss key documents, or provide context the model misinterprets.