Microsoft IQ
How Foundry IQ Connects to Foundry
Foundry IQ is the knowledge grounding layer for AI solutions built in Microsoft Foundry. It federates enterprise knowledge from multiple sources into a governed retrieval engine, giving agents accurate, permission-aware, and citation-capable answers.
What is Foundry IQ?
Foundry IQ is a fully managed knowledge system, currently in preview, built on Azure AI Search and integrated with Microsoft Foundry (formerly Azure AI Foundry). It provides AI agents and applications with a single endpoint for high-quality, governed organizational knowledge.
Where traditional retrieval-augmented generation (RAG) implementations require organizations to build and maintain their own search indexes, embedding pipelines, and permission enforcement, Foundry IQ provides these capabilities as a managed service with enterprise governance built in.
How Foundry IQ handles grounding
Grounding is what makes AI answers factual instead of fabricated. Foundry IQ grounds agents by connecting them to real enterprise knowledge through several mechanisms:
Federated knowledge retrieval
Foundry IQ connects to Microsoft 365, SharePoint, Fabric IQ, OneLake, Azure Blob Storage, Azure AI Search indexes, the web, and (in private preview) MCP sources. This means agents can draw from your entire knowledge estate through a single retrieval system.
Permission-aware access
Foundry IQ respects Entra ID-based governance and user permissions from connected knowledge sources. Data classifications and sensitivity labels from Microsoft Purview are maintained through the indexing and retrieval pipeline. Agents only surface knowledge that users are authorized to see.
Multi-hop reasoning
Unlike simple vector search, Foundry IQ supports iterative search and multi-source reasoning. Agents can follow chains of information across multiple documents and data sources to assemble accurate, comprehensive answers.
Source citation
Foundry IQ enables agents to cite their sources, which reduces hallucination risk and gives users the ability to verify answers. This is essential for enterprise trust and compliance.
Where it relates to enterprise knowledge sources
Most organizations have knowledge spread across many systems. Foundry IQ's value is that it federates retrieval across these sources without requiring organizations to consolidate everything into a single repository. Common sources include SharePoint document libraries, OneDrive business files, Azure Blob Storage containers, existing Azure AI Search indexes, Fabric IQ ontologies and semantic models, and web content.
This federated approach means organizations can start with the knowledge sources that matter most and expand over time, rather than requiring a complete knowledge consolidation project before any AI agent can provide useful answers.
What buyers tend to misunderstand
Foundry IQ is not just search
Basic Azure AI Search provides indexing and retrieval. Foundry IQ adds enterprise governance, permission enforcement, sensitivity label awareness, multi-source federation, and managed retrieval quality. The difference matters for production AI deployments.
Foundry IQ does not replace Fabric IQ
Foundry IQ handles unstructured knowledge retrieval. Fabric IQ handles structured data semantics. An agent that needs to understand both policy documents and business metrics needs both layers.
Preview status means careful evaluation
Foundry IQ is currently in preview. Organizations should evaluate it for fit and plan for evolution, but should not build production-critical systems on preview capabilities without understanding the maturity roadmap.
Custom RAG may still be needed
Foundry IQ covers many common grounding scenarios, but organizations with specialized retrieval requirements, unusual data formats, or non-Microsoft knowledge sources may still need custom RAG implementations alongside or instead of Foundry IQ.
When advisory help is needed
Foundry IQ advisory is most valuable when organizations are deciding how to ground their AI agents, evaluating whether managed retrieval meets their governance requirements, or planning the architecture for multi-source knowledge access. The decisions made at this stage determine whether AI solutions produce accurate, governed answers or become another source of unreliable information.
Part of a larger system
Why these questions matter beyond Microsoft
Understanding Microsoft's intelligence layers is the fourth stage of the AI Advantage Framework. The first three stages address which work to fund, whether the information is usable, and whether the workflows can execute. Platform decisions make more sense when those foundations are in place.
AI Advantage Framework
AI Fit & Governance → AI-Ready Data → Operational AI → Microsoft Intelligence
Building AI solutions that need enterprise knowledge?
Foundry IQ is one piece of the grounding puzzle. Marquee Insights helps organizations decide what their AI agents need to know and how to provide it safely.