The Governed Data Layer That's Powering Anaqua's AI Transformation

Anaqua provides intellectual property management software and services to corporations and law firms worldwide, helping them manage patents, trademarks, and other intellectual assets. Headquartered in Boston with roughly 950 employees and backed by private equity, the company serves some of the most complex IP portfolios in the world. Like many private equity (PE)-backed businesses in 2026, Anaqua faced a clear directive from its investors: build a credible AI transformation story by end of 2026.

CIO Erik Bailey was named AI Transformation Lead alongside his existing role to make that happen. But his ambition went further: His goal was to make AI a core operating capability across every function of the business. 

Five systems, two AI tools, and no unified view

Anaqua was already running Microsoft Copilot for a significant portion of its workforce and had introduced Claude for analytics and engineering. This setup determined one of the first requirements: the architecture had to be AI-agnostic, able to connect any model, today or in the future, without rebuilding.

The second aspect to consider: the AI tools needed live access to the systems where Anaqua's actual work lived; Salesforce, NetSuite, Zendesk, Azure DevOps, and BambooHR. A centralized approach was the only way to ensure every team worked from the same governed layer. One that wouldn't break the moment a system or model changed.

Erik and his team explored several options before landing on Connect AI, but each hit a wall quickly:

• Native Microsoft Copilot connector for Zendesk ran with admin-level permissions, meaning support tickets restricted to the U.S. team were visible to European users. A data governance failure that disqualified it immediately.

• Open-source BambooHR connectors lacked the semantic depth needed for AI to act on the data meaningfully.

What Erik ultimately wanted wasn’t another point solution but a single control plane. One layer that could manage connectivity, governance, and context consistently, across every system and every AI tool the company used or might use in the future.

Building the control plane: One layer for every system, user, and agent

Over a six-to-eight-week evaluation, Anaqua put Connect AI through its paces against all five target systems. What stood out was how fast the team could go from connection to useful AI answer. Salesforce, NetSuite, Zendesk, and Azure DevOps were up and running without custom development.

"Even over the course of the trial, we were able to enhance the capabilities of Connect AI in ways that made it more workable for us. That velocity gave us confidence that this wasn't just a static platform but one that would actually grow with us."

— Erik Bailey, CIO, Anaqua

On the governance and security side, per-user authentication was built in from the start, meaning each query ran as the individual making it, not as a shared admin account. For Salesforce, Anaqua ran two connectors simultaneously: one per-user OAuth connection for staff with Salesforce licenses, and one shared service account for executives who needed read access but weren’t licensed users. The same principle extended to write access.

When Anaqua's professional services team asked to test write access to Certinia, a Professional Services Automation (PSA) tool built on Salesforce, Erik was comfortable granting it because Connect AI provides the governance layer on top. With Connect AI, he could assign write access to two users only, regardless of what permissions they held in the underlying system. Everyone else stayed read-only. 

Early results and what's next: One control plane for all AI agents

What Anaqua built is a "centralized governance, distributed implementation" model. IT controls the connections and the access layer; functional teams build on top of it themselves. With this control plane in place, they moved quickly from pilot to company-wide rollout with five use cases:

• IT and operations: Running a self-improving AI agent that monitors Connect AI query logs, identifies SQL failures, and generates updated Skills automatically, delivering measurable improvements across every connected system.

• Support team: Using AI to identify patterns in support tickets and get instant answers from Zendesk in ways the IT team never anticipated, with the business team owning and expanding the use case themselves.

• Professional services: Testing AI-assisted write access to Certinia, the project and billing system built on Salesforce, with access controlled at the Connect AI layer to prevent unintended AI writes.

• CEO and CFO: Asking Claude questions about AI tool run rates and consumption across Copilot and Claude to manage spend and model selection.

• Customer success: Pulling correlated data across Salesforce, Zendesk, and NetSuite to build cross-system client views.

"They're not even thinking about Connect AI. They're just saying, ‘This is Claude’ or ‘This is Copilot.’ It's the plumbing behind the scenes that's enabling it. And the best plumbing is invisible."

— Erik Bailey, CIO, Anaqua

Alen Amini, who leads IT operations, has a clear vision for where this is headed: "We want every employee to have secured, governed access to AI, every high-value workflow AI-enabled, and every use case measurable against business outcomes."

For organizations at the beginning of this journey, Erik's advice is straightforward:

"Think carefully about what you're trying to do. Plan it out first and then give it a try. Don't be afraid to fail! Fail fast if you're going to fail. What you're going to find is that what rapidly started as a trial becomes real. And that's where the excitement comes."

— Erik Bailey, CIO, Anaqua

CData Connect AI: Governed data access for enterprise AI

CData Connect AI gives AI models, Claude, Copilot, or any other governed, per-user access to live enterprise data across every system that matters. One platform. Any model.

Start your AI transformation with Connect AI today at cdata.com/connect-ai.