We started CData with a simple belief that enterprise data should be accessible securely, reliably, and without months of custom integration. No middleware mazes. No brittle point-to-point builds. Just a clean, governed connection between the systems you run your business on and the tools that power decisions.
That belief hasn't changed. But the world around it has.
AI is reshaping enterprise software faster than any shift we’ve seen. Gartner projects $2.5 trillion in worldwide AI spending this year alone. But spending isn't the same as shipping. The vast majority of generative AI initiatives still stall before they ever reach production. Not because the models aren't capable, but because the data infrastructure underneath isn't ready.
Our own State of AI Data Connectivity Report puts a number on it: only 6% of organizations consider their data infrastructure fully ready for AI. Six percent. Meanwhile, 71% of AI teams spend over a quarter of their implementation time just wiring data plumbing.
That's the gap we're closing. And today marks a new day for CData—a new brand, a major platform release, and a clear declaration of where we're headed.
A new brand for a new era
When we founded CData, we were a connectivity company. And we still are. But with the rise of AI, connectivity has taken on new meaning, and the role of connectivity has taken on a new level of importance. We've stepped into the gap between AI models and AI reality, as the first-of-its-kind data layer between AI and the business outcomes it's supposed to deliver—to provide the connectivity, context, and control that production AI requires.
Our new brand reflects that evolution. CData isn't just keeping pace with the AI era, but we're building the infrastructure it runs on.
See our new brand and what's behind it →
Connect AI: built for production, not demos
AI agents are only as effective as the tools they can access, the data behind them, an accurate business context, and the controls governing both. Today we're announcing major enhancements to CData Connect AI that address exactly this.
Connectivity everywhere. Live, read-write access to 350+ business systems, including on-premise data behind the firewall with our new Connect Gateway. Your AI operates against enterprise systems with governed access and granular control.
Context that actually matters. We've addressed the issue of context overload with a layered approach to tools, starting with Universal Tools (that allow read-only exploration across the entire dataset in a consistent way across all data sources), Source Specific Tools (tools designed to bring the most common operations within a source), and Custom Tools (giving IT teams the ability to easily design tools for specific operations). Together, these tools allow teams to provision the right tools based on the use case.
Control you can trust. Per-user authentication, native source-system permissions, full audit trails plus new SCIM 2.0 support and Custom OAuth Applications for enterprise-grade identity governance. Every query is authenticated, authorized, and auditable.
Read the full product announcement →
The 25% accuracy gap
As enterprises evaluate MCP providers, one thing has become clear: architecture makes a measurable difference.
We tested five MCP providers across 378 real-world enterprise queries across CRMs, project management systems, ERPs, data warehouses etc., and scored every response against ground truth.
CData Connect AI achieved 98.5% accuracy. Other providers struggled particularly with multi-step logic, relative dates, and write operations: the exact scenarios that define real enterprise workflows.
Why does this matter so much? Because inaccuracy compounds. And in enterprise workflows, even small inaccuracies compound quickly. Over multi-step processes, that can mean the difference between automation and operational risk.
We built Connect AI differently. Instead of translating natural language directly into API calls, we use a relational abstraction layer with source-level semantic intelligence that understands entity relationships, business conventions, and workflow rules. Architecture matters. The results prove it.
Download the full accuracy whitepaper →
Meet us at Gartner Data & Analytics Summit
We're at the Gartner Data & Analytics Summit this week—Booth #308. Come see the new brand, the new platform, and the data behind these accuracy benchmarks in person.
And don't miss our speaking session: "AI Agents and the Future of Digital Work with Microsoft." Our Chief Product Officer Ken Yagen joins Microsoft's Partner Director of Product Management, James Oleinik on Wednesday, March 11 at 11:15 a.m. EDT to lay out a blueprint for moving from AI pilots to production-ready agentic AI.
Learn more about our Gartner presence →
What comes next
I've spent my career building data infrastructure. The problems have evolved, but the core challenge hasn't: getting the right data to the right place at the right time, securely and reliably.
What's different now are the stakes. AI agents aren't just reading dashboards; they're making decisions, taking actions, and operating autonomously inside your business. The data layer powering them has to be accurate, governed, and built for production.
This is an inflection point for our customers and the exact right moment for CData. The AI problem enterprises need solved most is the one we've spent our entire existence mastering: making enterprise data accessible, contextual, and trustworthy.
A new day. What began as a vision for simpler connectivity has evolved into infrastructure for agentic AI. Let's build.
— Amit
