For decades, the application was king. You moved data to it, processed it there, then moved it somewhere else. The architecture was built around software at the centre, with storage as the obedient plumbing beneath it.
That model, said Matthew Oostveen, chief technology officer and vice-president for Asia-Pacific at Everpure, is now obsolete.
“Now we have the data at the centre, and we’re moving our storage, our processing, and our applications, around it, closer to where that data resides,” he noted.
This inversion, from application-centric to data-centric architecture, is the most significant new concept from Everpure since the enterprise flash storage company rebranded from Pure Storage earlier this year.
Oostveen spoke to Techgoondu in a media briefing ahead of its Accelerate conference held last week in Las Vegas, where the company unveiled what it called Universal Data Intelligence. This involves a software layer designed to discover, classify, and contextualise data across an enterprise’s entire estate, regardless of where that data sits.
The shift matters because of what Oostveen called data gravity. As organisations accumulate data from analytics workloads, edge locations, and AI pipelines, moving it becomes progressively more expensive and error-prone, he noted.
The traditional response of giving each application its own data ecosystem eventually produces silos. Every new application added to the estate creates another one.
A unified data plane collapses that structure. Instead of applications holding private copies of data, they draw from a shared, centralised layer. Integration becomes simpler.
Semantic context, the business meaning of data, can then be attached once and inherited everywhere. And AI models training or running inference against that data are not working from fragmented or stale copies.
In his conversations with chief information officers (CIOs), Oostveen said that they are not even aware of the full extent of their infrastructure estate.
“They’re not really sure how many servers they’ve got, how much storage they have, how many devices are connected to their network,” he noted. “And if that is true, then it is also true that they’re not aware of all the data inside their organisation.”
Three layers of intelligence
Universal Data Intelligence operates across three functions. It enables automated discovery of data across all storage whether on-premises, cloud, Software-as-a-Service (SaaS) platforms, and even IBM mainframes which have historically been opaque to third-party tools.
It can also automatically classify the data by sensitivity, for example, business-critical, private, personally identifiable, and tagged with semantic metadata that connects it to the business processes it supports.
This wrapping is like a “semantic knowledge graph” around the data estate, explained Oostveen. The software understands what the data means and how it should be used.
The last benefit is compliance automation. Regulation-to-data mapping is applied automatically based on classification, enabling organisations to respond to regulatory requests, produce audit-ready reports, and track jurisdictional data residency without maintaining a static inventory that ages between updates.
For enterprises that operate across many nations, the compliance burden is substantial. Banking CIOs in Singapore, for example, are managing data governance conversations daily, not weekly.
“If they operate across multiple jurisdictions, that’s a really big part of the job,” said Oostveen. Many of these regulations are also not static as they are amended and supplemented continuously, meaning compliance is a moving target that a manually maintained data map cannot reliably track, he noted.
Moving to software
The name change from Pure Storage to Everpure signals a strategic repositioning that is now backed by product. The company still manufactures what Oostveen describes, with characteristic understatement, as “the greatest, fastest systems in the world”, including a new FlashArray//XL190 announced at Accelerate.
But the Universal Data Intelligence is sold as software, a storage-vendor agnostic solution that can scan competitor arrays as readily as Everpure’s own.
“We’re not here to compete with Snowflake, not here to compete with Splunk,” Oostveen said of the data and analytics vendors. “But we’re here to work with those types of data analytical tools to make it easier for the processing of information, and really to stay on the right side of regulation compliance.”
The positioning is deliberately layered above the hardware conversation. What Everpure is selling to CIOs and chief data officers is not a faster array but the ability to answer questions they currently cannot – what sensitive data do we hold, where does it live, is it compliant, and is it safe to use in an AI model?
That the company has chosen to make the software storage-agnostic broadens the addressable market considerably, while the integrations with its own Purity operating system offer a performance incentive for customers already in the Everpure estate.
Everpure’s new capability rests on an acquisition. In February 2026, Everpure bought 1touch, a contextual data intelligence startup whose technology forms the engine beneath Universal Data Intelligence.
The 1touch acquisition brought not just data discovery but also added inference-based classification. This is the ability to understand how data is actually being used across business processes, not merely where it sits.
Universal Data Intelligence is available now, with some customers already running it. The first deployments are probably in government. One application is in citizen services environments where the capability to correlate identity records across multiple data sets supports border security operations.
