Summary
Tip
See the Text and images tab for more details!
In this module, you learned how Microsoft Foundry IQ provides a managed, reusable knowledge layer for AI agents and applications. Foundry IQ uses Azure AI Search as its underlying indexing and retrieval infrastructure.
You explored how a knowledge base organizes one or more indexed or remote knowledge sources. These sources can connect enterprise content in systems such as Azure Blob Storage, SharePoint, OneLake, and Azure AI Search, as well as public web data. You also learned why content quality, freshness, metadata, identities, and source permissions are important to reliable retrieval.
Foundry IQ uses agentic retrieval to plan searches, decompose complex questions, search relevant sources, rank results, and return evidence with citations. Agents can connect to a knowledge base through an MCP tool and use that evidence to generate grounded responses. Clear instructions, representative testing, and separate evaluation of retrieval and generation help make the experience more trustworthy.
For more information, see Foundry IQ