73% of organizations cite data connectivity—not models—as their primary barrier to scaling AI. The model isn't the bottleneck. The plumbing is.
At the upcoming Gartner Data & Analytics Summit, CData and Microsoft are laying out a blueprint for what comes next: AI agents that work as colleagues, not just tools.
Register now
The shift from pilots to production
Gartner's message is clear: 2025 was about experimentation; 2026 is about delivering agentic AI ROI. But most organizations aren't ready. Shadow AI is spreading, governance confidence is low, and Gartner predicts more than 40% of agentic AI projects will be canceled by 2027.
The gap isn't in the models—it's in the architecture.
Three gaps blocking governed AI at scale
Enterprise leaders are investing to close three critical gaps:
Connectivity — AI can't reach the data it needs, forcing incomplete decisions or risky workarounds.
Context — AI can't understand business meaning, leading to inaccurate results or wrong actions.
Control — AI can't be governed reliably, with no guardrails, no audit trail, and no way to enforce policy at the data layer.
CData's research confirms enterprise leaders know this—their top investment priorities are Control (60%), Connectivity (42%), and Context (34%). Leaders are 2x more likely to invest in data infrastructure than model development.
The partnership: Microsoft + CData
Together, Microsoft and CData address all three. Microsoft delivers the intelligence layer through Work IQ, Copilot Studio, and Agent 365. CData extends that reach with 350+ enterprise data source connectors, semantic intelligence so agents understand business meaning, and inherited identity with full audit trails.
The result: 98.5% accuracy compared to 65-75% for alternative MCP providers—the difference between copilot-ready and autonomous-agent-ready.
One architecture for the entire spectrum
The future of work isn't a binary switch from human to agent. It's a spectrum—from humans exploring data, to agents executing workflows, to agents reasoning autonomously. The organizations that win will build architecture that supports that entire spectrum through the same governed infrastructure.
Join us at Gartner to see this in action: a live demo of cross-functional agents orchestrating order-to-cash processes across Salesforce and Dynamics 365, with inherited RBAC, full audit trails, and workspace isolation—security built in, not bolted on.
Meet the speakers
Ken Yagen, Chief Product Officer, CData Ken brings more than 25 years of experience driving innovation in enterprise software. His career includes product leadership roles at MuleSoft, where he shaped strategy for APIs and integration platforms, as well as positions at Box and Symphony. Ken is leading CData's product strategy as the company scales its AI connectivity platform for enterprises deploying agentic AI.
James Oleinik, Partner Director of Product Management, Microsoft James is a Partner Director of Product Management at Microsoft, specializing in AI and analytics for low-code platforms. He leads product management for Microsoft Dataverse, AI Builder, and Power Platform Connectors—the foundation powering Microsoft Copilot Studio and Agent 365. He holds a Master of Engineering and a Bachelor of Science from the Massachusetts Institute of Technology.
