Informatica and Microsoft expand partnership for agentic AI built on governed data
Informatica and Microsoft are tightening their integration to make agentic AI deployable on trusted, governed data. The move connects Informatica's Intelligent Data Management Cloud (IDMC) with Microsoft's AI Foundry, enabling AI agents to tap directly into catalog, data quality, and master data management services. A new Model Context Protocol (MCP) integration lets agents built in Azure AI Foundry discover and access relevant enterprise datasets in near real-time. The goal: faster builds, cleaner access patterns, and stricter compliance for production use.
What's new
- MCP integration between Azure AI Foundry and Informatica IDMC for near real-time data access.
- Direct hooks into Informatica's catalog, data quality, and MDM services for trustworthy agent responses.
- Enterprise-grade interoperability designed to respect existing security and compliance controls.
GenAI recipes that cut build time
Informatica is rolling out pre-built "GenAI recipes" to speed up common patterns: retrieval-augmented generation (RAG), chat history handling, and prompt chaining. These templates combine Azure OpenAI tools with Informatica's data services to reduce glue work and policy drift. Initial use cases include loan processing and automobile insurance claims handling. Informatica's CLAIRE AI engine is now natively available across Azure regions in the US and Europe, helping teams meet regional data handling standards.
Why this matters for IT and development teams
- Shorter path from proof-of-concept to production with standardized templates and governed data access.
- Consistent data controls across agents: lineage, quality checks, and master data provide higher signal-to-noise.
- MCP reduces custom connectors and one-off plumbing that increase risk and maintenance.
How to put it to work
- Use Azure AI Foundry to build agents that call IDMC via MCP for data retrieval, validation, and enrichment.
- Scope agent access to approved datasets through the catalog; apply data quality checks before generation.
- Adopt the RAG and prompt-chaining recipes for your first workloads, then tune prompts and evaluation loops.
- Log agent actions and responses; review for drift, data leakage, and compliance exceptions before scaling.
Analytics expansion: OneLake + Apache Iceberg
Informatica is extending support for Microsoft OneLake tables backed by Apache Iceberg. This helps teams unify data across sources while staying aligned to open table formats and analytics standards. Iceberg compatibility makes it easier to operate multi-engine analytics without brittle conversions.
Customer and partner perspective
"Our collaboration with Microsoft brings the best of Microsoft Foundry and enterprise-grade data management together. By integrating our CLAIRE AI engine and IDMC services with Foundry, we help customers build AI agents and applications with confidence, compliance and speed," said Krish Vitaldevara, Chief Product Officer, Informatica.
Dayforce is already seeing traction. "Our partnership with Informatica helps Dayforce develop a unified, reliable view of our customers - the foundation for providing smarter, more personalised experiences," said Anupama Jha, Vice President of Enterprise Data Governance and Analytics, Dayforce.
Microsoft echoed the focus on trustworthy data. "Enterprises are adopting agentic AI through Microsoft Foundry. Trusted, governed data is critical to delivering accurate and responsible outcomes. Informatica's deep integration with our platform empowers customers with the data they need to bring AI to life in production," said Amanda Silver, CVP Apps and Agents, Microsoft.
Practical next steps
- Inventory high-value datasets in Informatica's catalog and tag what's safe for agent access.
- Start with a narrow RAG use case, wire in data quality checks, and measure response accuracy and latency.
- Stand up observability for prompts, tool calls, and data access paths; review regularly with security and compliance.
- Pilot Iceberg-backed OneLake tables where you need multi-engine analytics with consistent table semantics.
Learn more about the platforms involved: Azure AI Foundry and Apache Iceberg.
If your team is building agentic workflows and needs structured upskilling, explore focused tracks for developers and IT: AI courses by job.
Your membership also unlocks:
