In a documentation context, federated search means a reader on any one property, such as the main docs site, an API reference, or a knowledge base, can search once and get answers drawn from all of them. The content stays where it is published. The search layer indexes each property and merges the results.
This matters because documentation rarely lives in one place. A typical product has a docs site, a changelog, tutorials, and a support knowledge base on different platforms, and a reader's question does not respect those boundaries. Without federated search, each property only answers from its own content, and questions that cross a boundary dead-end.
An AI layer extends federated search from ranked results to direct answers: the same combined index can power a chatbot that synthesizes an answer across properties and cites the pages it drew from, wherever they live.