Knowledge base
The material an AI assistant answers from: price lists, terms, FAQs and pages on the website.
Knowledge base
RAG.
Retrieval-augmented generation. A technique where the AI first fetches relevant parts of the material and then answers from them.
01Explanation
How it works
Instead of answering from what the model learnt during training, the system first searches a knowledge base for the passages that fit the question. They are passed to the language model, which phrases the answer from them.
This lets the answers rest on up-to-date details from a particular company, and lets them point to where the answer comes from. It reduces the risk of invented answers but does not remove it entirely.
02In practice
In customer service
Assistants that answer from a company's own material are often built on some form of RAG. That is why an answer can change the same day the terms change, without any model having to be retrained.
04Get started
See how it works for you
In a short demo we show how an AI assistant can answer your customers' questions, using your own material. It takes a quarter of an hour. From SEK 3,900 a month excl. VAT, with the exact price set after the demo.