At its Accelerate user conference, Everpure expanded its platform with an organizationally separate data intelligence business and brought several storage products to general availability.
At its Accelerate user conference in Las Vegas, storage vendor Everpure (NYSE: P) unveiled a sweeping overhaul of its Enterprise Data Cloud platform. At the core of the announcement is a strategic pivot: rather than organizing its architecture around applications, the company is rebuilding it around the data itself. The driver is the growing use of AI agents, which, unlike conventional software, need to reach across system boundaries rather than settling for whatever slice of data a single application happens to hold.
Under the application-centric model that has shaped enterprise IT for decades, information lived wherever the relevant business application stored it. Every time data moved into a new context, another copy was created, and ETL, MDM, and BI tooling existed largely to reconcile the resulting mess after the fact. According to Everpure, that model no longer holds up for AI agents, which require a current, cross-system view of enterprise data rather than the partial picture any one application can offer.
The rebuild rests on three components: a data intelligence layer, a unified storage plane, and an intelligent control plane that governs the entire infrastructure fleet.
The most notable move is the renaming of 1touch.io, a company Everpure acquired some time ago, into Everpure Data Intelligence. The tool discovers, classifies, and contextualizes enterprise data wherever it sits, drawing on more than 300 database connectors alongside file shares and S3 buckets. It maps raw data to its business meaning and, in the process, supports compliance reporting obligations such as those under GDPR. Notably, Everpure is running this unit organizationally apart from its core storage business, positioning data intelligence to grow as a separate line with its own classification logic and customer-specific policies.
On the storage side, Everpure is extending the FlashArray//XL 190 with a Purity Turbo mode designed to wring more performance out of existing hardware, aimed in particular at read workloads that currently oversubscribe a single array. With Everpure Data Stream, now generally available, the company is simplifying the construction of data pipelines for retrieval-augmented generation: extraction and transformation of unstructured data run on Nvidia’s software stack, with scaling that spans from FlashBlade//S up to exabyte-scale FlashBlade//EXA deployments. General availability of Everpure Cloud Azure Native Virtual Machines, manageable like a native Azure service, is expected to follow in July. The lineup is rounded out by Evergreen//One Overdrive, which lets customers set their own performance and capacity ceilings, exceed them temporarily when needed, and pay only for what they actually use — a cost-versus-risk trade-off the company says will become available starting in the third quarter.
The third pillar, the Intelligent Control Plane, aims to make hundreds of storage systems manageable through a single interface, API, or command line, addressing a common customer complaint about juggling too many applications, platforms, and data silos at once. It bundles automated compliance monitoring, workload rebalancing across systems without downtime, and agentic workflows built on the Model Context Protocol. What remains unresolved is how far governance and security for autonomously acting agents can actually be automated — a challenge Everpure itself describes as one of the hardest unsolved pieces of this overhaul.
A shift in go-to-market strategy is also worth noting. The new data intelligence business reportedly reaches a largely different set of buyers than the traditional storage business and leans more heavily on indirect channels. In Germany, that means established integrators such as SVA, Computacenter, and Bechtle alongside newer boutique partners are expected to play a larger role — a sign that the vendor sees this new line as requiring different sales expertise than conventional storage infrastructure.
Alongside the product news, Everpure is launching the EDC Success Blueprint in July, a three-stage methodology combining a self-assessment, capability-specific guides, and a facilitated workshop. It is meant to help customers locate their own maturity across ten categories, from cloud agility to cyber resilience, and turn that assessment into a migration roadmap. That matters for existing customers because shifting from an application-centric to a data-centric architecture is not purely a technical change; it also redraws responsibilities across IT, security, and business units.
For organizations weighing whether to build their AI strategy on this architecture, the underlying question Everpure itself raises in its materials remains open: large language models are not automatically suited to every enterprise context, and the quality of the underlying data matters more to outcomes than model size. Before any project can scale from small language models to enterprise-wide AI factories, companies first need basic visibility into what data they hold, where it sits, and how it relates — an inventory that, in many IT departments, has yet to be completed.
Dr. Jakob Jung is Editor-in-Chief of Security Storage and Channel Germany. He has been working in IT journalism for more than 20 years. His career includes Computer Reseller News, Heise Resale, Informationweek, Techtarget (storage and data center) and ChannelBiz. He also freelances for numerous IT publications, including Computerwoche, Channelpartner, IT-Business, Storage-Insider and ZDnet. His main topics are channel, storage, security, data center, ERP and CRM.
Contact via Mail: jakob.jung@security-storage-und-channel-germany.de