Glossary
What is Retrieval-augmented generation?
Retrieval-augmented generation is a method where an AI looks up relevant text from your own documents before it writes an answer.
How it works
Your documents are split into small pieces and indexed. When a question arrives, the system finds the pieces most relevant to it. It passes those pieces to the model with the question, and the model writes an answer from them.
Why businesses use it
It keeps answers close to your current policies, prices and product details. You update a document, and the next answer reflects the change. It also makes it easier to check where an answer came from.
Limits to know
Poor source documents give poor answers. Retrieval can miss the right passage. Old or contradictory documents confuse the result. Keep the knowledge base clean and dated.
Example
A Bengaluru school puts its fee schedule, admission rules and calendar into a knowledge base. A parent asks whether a sibling discount applies. The system finds the fee policy passage, and the answer quotes the rule from that policy.
Montr’s chatbot, which is in beta, answers from the Documents and forms you point it at. The Inbox also has a Knowledge base that the support bot and AI suggestions draw on.