CData is moving from being the company that connects data to the company that teaches AI how to understand it.
In late September 2025, CData introduced Connect AI, described as the first managed Model Context Protocol (MCP) platform built to give AI applications governed, live access to enterprise systems. Connect AI is designed to help AI assistants and agents use enterprise data in real time while inheriting permissions and security protocols of the source systems.
Connect AI extends the connectivity layer CData has spent years building—now spanning more than 350 enterprise systems—into AI-native workflows. Instead of extracting data into separate pipelines, the platform connects AI agents directly to live systems, preserving user permissions, metadata, and semantic relationships. In practical terms, that means agents can reason across Salesforce, SAP, NetSuite, Snowflake, and other systems without custom integration work or brittle scripts.
The launch quickly expanded beyond CData’s own ecosystem. In November, Connect AI became available in the Databricks Marketplace, enabling Agent Bricks users to deploy AI agents with real-time access to enterprise systems. Just two weeks later, CData announced integration with Microsoft Copilot Studio and Microsoft Agent 365, embedding MCP connectivity into Microsoft’s agent framework. Together, those integrations place Connect AI closer to the platforms many enterprises already use to build and deploy agents.
CData also leveraged that same stretch to make a product leadership hire. In late November, the company appointed Ken Yagen as Chief Product Officer to lead product strategy and engineering. With a background that includes MuleSoft, Yagen joins as CData expands Connect AI and its MCP tooling for both enterprise teams building agents internally and software providers embedding agent capabilities into their products.
The timing reflects a broader reality in enterprise AI. Billions are being spent on pilots, yet production deployment often stalls at the same bottleneck: governed access to live, contextual data. CData’s bet is that MCP-based connectivity—managed, secure, and semantic-aware—becomes the connective tissue that allows agents to move from experimentation to operational use.
If the last decade was about making data accessible, CData’s recent stretch suggests the next one is about making that data usable by systems that think and act on behalf of the enterprise.