Enterprises racing to deploy AI are discovering that the bottleneck is not compute or models, but rather the failure to adopt a data-centric model to resolve unmanaged, ungoverned data sitting across silos that can’t be classified, accessed or trusted at production speed.

That pressure is precisely what drove Everpure Inc.’s rebranding from Pure Storage and its pivot toward a data-centric model built around intelligence and governance. The company is no longer positioning itself as a storage provider that happens to run AI workloads; it’s staking out the full data layer, according to Lynn Lucas (pictured, left), chief marketing officer of Everpure.

“The business strategy to move and expand into data management, and now Data Intelligence, we felt might not be as resonant with a company name with ‘storage’ in it for the new audience that we are expanding into,” she said. “The newer folks that we are addressing, chief data officers, chief AI officers … really felt like ‘storage’ in our name might be limiting.”

Lucas and Phil Goodwin (right), research vice president of multicloud data management and protection at IDC Corp., spoke with theCUBE’s Christophe Bertrand and Alison Kosik at Pure Accelerate 2026, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the company’s rationale for rebranding, Everpure’s shift toward a data-centric model, and the governance gap that prevents enterprises from realizing AI return on investment. (* Disclosure below.)

A data-centric model addresses governance barriers to AI success

The rebrand reflects a real market shift that IDC research is tracking closely. Enterprise AI failures are rooted in data dysfunction, not model limitations, Goodwin noted.

“This pivot to more of a data-centric model rather than hardware centric, I think, is very well timed,” he said. “The research that we’ve done has shown that the real barriers to AI success start with governance. In fact, that’s the number one reason that AI projects fail. The second one is data access — the silos, the inability to bring in the data.”

About 54% of AI projects never make it to production, representing zero return on investment for companies that have already committed significant capital, Goodwin noted. The answer is not simply pooling data; it’s governing what goes where. Enterprises that treat governance as an afterthought, rather than a foundational control layer, are the ones left with failed pilots and mounting costs.

“Not all data belongs in every learning module,” Goodwin said. “You really need to be able to differentiate what data needs to go where, how it needs to be protected, how it needs to be governed. Maybe it’s sovereign data … it’s really that data control layer that’s critical in an AI environment.”

Everpure Data Intelligence, built on the company’s acquisition of 1touch.io Inc., is designed to deliver that control regardless of where data lives — on-premises, in the cloud or outside of Everpure’s arrays, Lucas noted. The vision is a kind of navigation system for enterprise data: not just mapping where data sits, but surfacing the context that makes it actionable and safe for AI.

“You need AI-ready data with your AI-ready infrastructure in order to get the business outcomes for your organization,” Lucas said. “Make sure you are working on both in order to deliver to the business what you need.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Pure Accelerate 2026 event:

(* Disclosure: TheCUBE is a paid media partner for Pure Accelerate 2026. Sponsors of theCUBE’s event coverage do not have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE