Las Vegas does subtlety about as well as it does quiet, so maybe it’s fitting that Everpure places a few bets of its own during Pure Accelerate 2026. And though the company announced its name change back in February of this year, that name change was still something people were talking about – in the audience and on stage as well.
Walk into the keynote hall expecting the usual array refresh and faster-flash victory lap, and you got something else entirely. The company is now Everpure, and the Day 1 keynote was less a product launch than an argument about how enterprise IT has been built wrong for 30 years.
Let me back up, because the name change is the easy headline and the smaller part of the story.
A new name, and the reason behind it
Chief Marketing Officer Lynn Lucas opened by naming the obvious. “You may notice something a little different this year. We are Everpure.” The explanation was straightforward: after a year of talking to customers and partners, the company kept hearing that it was viewed as more than a storage vendor, and the old name had started to box them in.
I asked Everpure’s Chief Technology and Growth Officer, Rob Lee, about this directly in our 1:1, because a rebrand is rarely just a logo. He was candid about the tension. “Pure Storage was a great name for an era of the company,” he told me, “but it did also start becoming a bit pigeonholing.” His phrase for the decision: they wanted to keep an element of “Pure” but felt it was “time to let storage go.”
The keeper detail is that this wasn’t a marketing whim. Lee described it as a multi-year, intentional move tied to the company’s expansion from storage infrastructure into full-stack data management. New name, new buyers, new conversations inside the same accounts.
The real keynote: flipping IT on its head
Everpure CEO Charlie Giancarlo took the stage and, by his own admission, surprised the room. “I’m not going to talk to you about data storage,” he said. “I want to talk to you about your data.”
His thesis goes like this. For three decades we’ve built IT around applications. ERP was supposed to be the one system to rule them all, then the workflow vendors showed up—CRM, Workday, ServiceNow, NetSuite—and each one fragmented the data a little more. The result is an enterprise where “customer” means something different in every system, teams burn hours reconciling definitions, and nobody can answer a basic question like “tell me everything about this customer” without a small data-integration project.
Then AI arrives and makes the whole mess worse. Every AI vendor, every SaaS vendor, every analytics tool wants a copy of all your data. Giancarlo’s blunt question to the room: “How many vendors are you going to give all your data to?”
His answer is what Everpure is calling data primacy. Instead of applications owning the data, you build a governed system of record for the things that actually define your business—for Everpure, that’s “customer”—and you create a shared context layer (think a semantic knowledge graph) that maps how data relates across systems. Apps, analytics, and AI agents then read from and write to that source of truth instead of hoarding their own copies. Fewer integrations, fewer copies, smaller attack surface, more coherent data.
Here’s the honest caveat Giancarlo himself put on it: this is a long-term vision, not a switch you flip. He told us during the press Q&A it’s roughly a two-and-a-half-year journey inside Everpure’s own IT, where they’re running it as an internal program code-named “Mercury” with a weekly meeting he personally attends.
The product news that makes it real
A vision is just a slide deck without shipping product, and Day 1 had plenty. Two press releases landed, anchored by the data-primacy story.
- Everpure Data Intelligence (the rebranded 1touch.io acquisition) discovers, classifies, and contextualizes data wherever it lives—on the Everpure platform, public clouds, SaaS apps, even mainframes and third-party storage. It builds the semantic knowledge graph that maps raw data to its real business meaning, flags PII and PHI, and tracks lineage. Available now.
- Data Stream, built on NVIDIA’s AI Data Platform reference design, automates the grind of getting unstructured data AI-ready—ingestion, curation, indexing, vectorizing—and aims to cut work that takes data engineers months down to minutes.
On the infrastructure side, Chad King and the demo team showed the Enterprise Data Cloud maturing into three layers: a Unified Data Plane that stores everything from archive to AI, an Intelligent Control Plane that governs it, and the new Universal Data Intelligence layer that understands it. The Fusion control plane got real teeth—synchronous replication extended to file via active cluster, fully automated workload rebalancing and mobility (coming Q4 2026), and Fusion compliance reporting with agentic triage. The standout demo, for my money, was the ransomware-policy scenario: change a snapshot retention baseline once, and Fusion enforces it across the entire fleet instead of you logging into dozens of arrays by hand.
There were dollar figures attached, too: Evergreen//One Overdrive (Q3 2026) for temporary on-prem performance bursts up to 25% over baseline, Azure native VM support for Everpure Cloud now generally available, and FlashBlade//EXA pushing past 7,200 on SPEC AI.
Charlie didn’t dodge the NAND elephant
You can’t talk about Day 1 honestly without the pricing moment, because Giancarlo didn’t skip it. NAND costs, he said, have jumped between 600% and 800% in six or seven months. Everpure’s cost to produce a product is up over 300%. The company raised prices—less than the cost increase—and is now running at the low end of its gross margin range to absorb some of the pain rather than pass it all on.
In the press Q&A, a reporter pressed him on a competitor accusing Everpure of profiteering. His rebuttal was the financials: operating at the low end of gross margin is the proof point, he argued, and you can check the quarterly numbers. This was a rare bit of straight talk in a keynote season usually allergic to the topic.
The customer who lived it
The Fiserv segment grounded the whole thing. CTO Pete Cavicchia (running infrastructure that powers billions of financial transactions and the Clover point-of-sale platform) described moving off the old “data locked in isolated silos” model toward Everpure’s Enterprise Data Cloud. His line that stuck: data is “no longer the cost-driving byproduct, it’s the revenue-driving product.”
What an IT pro should take away
When I asked Rob Lee for the single takeaway for VMblog readers, he didn’t point at a product. “Our medium-to-longer-term vision of how the IT architecture landscape is going to change,” he said, then sharpened it: “all of the traditional IT assumptions are getting turned on their heads.” Not AI-washing, he stressed—first principles.
So here’s my read after Day 1. The big story is whether data primacy is a genuine re-architecture or a very well-produced reframing of “data-centric,” an idea the industry has flirted with for a decade. Giancarlo’s own argument is that AI is the breaking point that finally forces it, and the proof points—Data Intelligence shipping now, Data Stream with NVIDIA, Fusion getting autonomous—suggest they’re serious. Looking forward to more 1:1 interviews and to the Day 2 keynote which will hopefully go deeper on the vision. I’ll be there, and I’ll be watching the Expo floor to see how 14-plus partners feel about a “storage” company that just declared the data layer its home turf.
##
