What is an LLM?

An LLM is a large language model: a program trained on a large amount of data that can, among other things, create, analyse and transform text. While generating, it picks the next pieces of the answer based on the input and what it learned earlier.

It can help write a description, extract data from a document or decide which tool to use. But it isn't automatically an up-to-date database of facts. It can give a convincing but untrue answer. That's why, for important data, we add verification, sources and limits on what it can do. A model call is one part of a system, not a complete application.