Informatica, now a Salesforce company, on May 20, 2026 announced deep new integrations with Microsoft that bring its Headless Intelligent Data Management Cloud into Microsoft Foundry via MCP servers and extend Microsoft Fabric with built-in change data capture (CDC) capabilities. The move aims to accelerate enterprise AI adoption by making trusted, governed data immediately accessible to AI agents and copilots within the Microsoft ecosystem.

Informatica's IDMC platform becomes one of the first major data management suites to leverage the Model Context Protocol (MCP) as a bridge between AI systems and enterprise data assets. MCP, an open standard originally pioneered by Anthropic, defines a universal interface for connecting large language models to external tools and data sources. By exposing IDMC's catalog, quality, and governance functions through MCP servers, Informatica allows AI agents in Microsoft Foundry to dynamically discover, request, and retrieve data with full transparency into its lineage, quality, and access policies.

For Microsoft customers, this means a significantly shorter path from AI ambition to production. Instead of manually coding point-to-point integrations, developers can now register Informatica's MCP servers as trusted endpoints within Foundry's agent-building surface. The agents then autonomously fetch datasets, apply quality checks, and ensure compliance with corporate policies—all without human intervention. Informatica's metadata-driven automation ensures that the context surrounding the data travels with it, enabling more accurate and responsible AI outputs.

How the MCP Server Integration Works

The integration centers on a set of purpose-built MCP servers that encapsulate IDMC's core capabilities: data cataloging, data quality, metadata management, and master data management. Each server exposes a standardized, JSON-RPC-based interface that AI agents can invoke through natural language commands. For example, a financial analyst building a forecasting agent in Foundry could simply ask, "Get me the latest sales data from the North America region, ensuring it matches the golden record definitions." The agent would use the MCP servers to query the IDMC catalog for the appropriate dataset, validate it against data quality rules, and return only the certifiably accurate rows.

This tight coupling eliminates the traditional friction between data engineering and AI teams. Data stewards define rules and policies once in IDMC, and they are automatically enforced when AI agents consume the data. Governance is no longer an afterthought; it is baked into the AI pipeline from the start. Microsoft Foundry's orchestration engine handles the execution flow, while Informatica's servers provide the data intelligence layer.

Microsoft Fabric CDC: Real-Time Data for AI Workloads

Simultaneously, Informatica expanded its partnership with Microsoft Fabric by introducing native Change Data Capture (CDC) capabilities. CDC, a technique for capturing real-time changes from transactional databases, is now available as a first-class feature within Fabric's Data Factory. Users can set up low-latency data pipelines that stream inserts, updates, and deletes from sources like SQL Server, Azure SQL Database, and Oracle directly into Fabric lakehouses or warehouses, with minimal configuration.

This enhancement is critical for AI scenarios that demand fresh data. A retail recommendation engine, for instance, needs near-instant awareness of inventory changes or purchase events. With Informatica's CDC engine integrated into Fabric, such pipelines can be built in minutes using a simple visual interface, without writing complex streaming code. The integration also respects the unified Fabric security model, meaning row-level security and data masking policies applied in Fabric extend to the CDC streams automatically.

A Unified Platform for Enterprise AI

The duo of MCP-based intelligent data access and real-time CDC addresses two of the biggest bottlenecks in enterprise AI: data trustworthiness and data timeliness. By embedding these capabilities directly into Microsoft's AI and analytics platforms, Informatica and Microsoft are creating a cohesive environment where data flows freely yet securely from operational systems to AI models, with full observability at every step.

Informatica's IDMC already integrates with over 200 data sources and provides more than 1,000 prebuilt transformations. Now, through Foundry, those assets become directly consumable by the next generation of AI applications. Early adopters in financial services and healthcare have reported cutting data preparation time by up to 70% while simultaneously reducing the risk of AI hallucinations caused by stale or unauthorized data.

Data Governance and Compliance Reinforced

Governance remains a central pillar of the announcement. Informatica's long-standing strength in policy management and data stewardship is now accessible to AI agents via the MCP servers. Every data request through the servers is logged, audited, and subject to the same controls as human-led access. This audit trail is vital for regulated industries where explainability is mandatory. If an AI agent makes a recommendation based on a dataset, a compliance officer can trace back exactly which rules were applied, who owns the data, and whether any transformations might have affected the outcome.

Microsoft Foundry's built-in responsible AI tooling further complements this by allowing developers to monitor agents for bias, fairness, and reliability. The combined solution effectively creates a closed loop from data ingestion (via CDC) to AI decision support (via Foundry agents) with governance checkpoints at every stage.

Competitive Landscape and Market Impact

The announcement positions Informatica as a key enabler for Microsoft's AI strategy, competing directly with other data integration vendors like Fivetran, Qlik, and Talend. However, the depth of the MCP integration sets it apart. While others offer connectors and APIs, Informatica's adoption of an open protocol standard reduces vendor lock-in and simplifies multi-agent coordination. As the MCP ecosystem grows, any AI agent that speaks MCP—whether from Microsoft, Anthropic, or open-source communities—can tap into IDMC's governed data fabric without bespoke integrations.

For Microsoft, this strengthens Foundry's appeal to enterprises that already standardized on Informatica for data management. It also alleviates a common pain point: the challenge of making on-premises and legacy data sources available to cloud AI services. With Informatica's wide connectivity and CDC features, even mainframe data can stream in real time to Fabric and be immediately used by AI agents.

Implementation and Availability

The MCP server integration is available immediately for Informatica customers with the Intelligent Data Management Cloud and an active Microsoft Foundry subscription. Initial deployment involves registering the servers through a simple configuration process in the Foundry portal, where administrators can set access scopes and rate limits. The Fabric CDC capability entered public preview on May 20, 2026, with general availability expected in Q3 2026. Pricing follows Informatica's consumption-based IPU (Intelligent Processing Unit) model, with special bundled offers for Microsoft enterprise agreement holders.

Thousands of organizations using both platforms can benefit immediately. Early adopters include a global bank using the integration to power real-time fraud detection agents, and a healthcare network building a patient data harmonization agent that adheres to HIPAA governance rules.

Looking Ahead

Informatica and Microsoft hinted at deeper synergies on the horizon, including MCP servers optimized for specific verticals (such as supply chain or customer 360) and tighter integration with Microsoft Copilot for Data in Fabric. As AI agents become the primary consumers of enterprise data, the partnership aims to make data quality, governance, and freshness as automatic as electricity—reliable, invisible, and essential.

The move signals a broader industry shift toward protocol-based data access for AI, moving away from fragile, custom-built pipelines. For Windows enthusiasts and Microsoft ecosystem architects, the message is clear: the data foundations must be just as intelligent as the AI that runs on them. Informatica and Microsoft are betting that together, they can provide exactly that.