Short answer
Retrieval-augmented generation, or RAG, is a way of building an AI system so that it looks up relevant passages in a set of documents and uses them when it writes an answer, instead of relying only on what the model learned in training. The idea comes from a 2020 research paper by Lewis and colleagues.
How it works in a team bot
When someone asks a question, the system first searches your approved answers for the passages that match, then asks the model to write the reply using only those passages. If nothing matches, a well built system says it does not know instead of guessing.
What it does and does not fix
Looking things up makes answers more specific to your documents. It does not guarantee the model uses the passages faithfully, and it is only as good as the documents behind it. Old or conflicting documents produce old or conflicting answers.
What to check
Ask where an answer came from and whether you can open the passage. Test with questions your documents do not answer, and see whether the system admits it.
Related
Want early access and new checklists by email?
Leave your email and we write to you when your spot is ready. Or sign in now and the first 100 founders start right away.
Common questions
No. Training changes the model. RAG leaves the model as it is and gives it relevant passages from your documents at the moment of the question.
It reduces them by grounding answers in your documents, but it does not remove them. Test the system with questions that have no answer.