Back

Everpure

Everpure (Formerly Pure Storage)

Everpure is the name Pure Storage adopted on February 23, 2026. The company makes all-flash storage systems, including FlashArray and FlashBlade, and sells them through subscription models such as Evergreen. It renamed itself to reflect a strategic shift from storage hardware to data management, positioning its platform as an Enterprise Data Cloud. At the same time it announced the acquisition of 1touch, a data discovery and classification company. The stock ticker, PSTG, did not change.
Source: Blocks and Files
Source: Everpure

What Changed and What Did Not

  1. Company name. Pure Storage became Everpure, and email and web domains moved to everpuredata.com.
  2. Product branding. Products are transitioning to Everpure branding through 2026. Portworx became Portworx by Everpure.
  3. Unchanged. The stock ticker, product functionality, existing contracts, SKUs, and support terms stayed the same.

The Everpure Portfolio

  1. FlashArray. All-flash block and unified storage for databases, virtualization, and file workloads.
  2. FlashBlade. All-flash scale-out file and object storage for unstructured data, analytics, and AI.
  3. Evergreen. Subscription and upgrade programs for Everpure storage.
  4. Fusion. A management layer that pools Everpure arrays across locations.
  5. Portworx by Everpure. Data services for Kubernetes and containerized applications.
  6. Everpure Data Intelligence, formerly 1touch. Data discovery, classification, and semantic context across data sets.

Everpure Data Intelligence and MCP

At Accelerate 2026, Everpure expanded data intelligence for production AI and made it available in October 2026. Everpure Data Intelligence, formerly 1touch, discovers, classifies and contextualizes data across Everpure arrays, other on-premises storage including mainframes, public clouds, and SaaS sources such as OneDrive, SharePoint and Google Workspace. It flags PII and PHI, tracks lineage and offers native Model Context Protocol (MCP) access to its data catalogs. Everpure also announced MCP servers for Fusion and Pure1 that let AI agents query infrastructure topology and performance metrics.

Source: Blocks and Files

Source: Computer Weekly

Everpure is one of several primary storage vendors, along with NetApp and Dell, now selling data intelligence and AI data preparation alongside arrays. The shift follows where the value has moved: unstructured data makes up 80% to 90% of enterprise data, it feeds AI, and it is expensive to keep on flash.

All-flash capacity is paying record prices. The top five enterprise SSD suppliers reported revenue of nearly $37.59 billion in the second quarter of 2026, up 103.6% from the first quarter, driven by higher contract prices. TrendForce projects NAND flash contract prices will rise another 15% to 20% quarter over quarter in the fourth quarter of 2026, with enterprise SSD the only memory category set to increase faster than it did the quarter before, and expects memory contract prices to remain broadly elevated in 2027.

Source: TrendForce, enterprise SSD revenue

Source: TrendForce, fourth-quarter 2026 forecast

Source: TrendForce, 2027 outlook

Gartner forecasts NAND flash revenue will grow 371.9% in 2026, driven largely by pricing, with supply and demand staying tight through 2027.

Source: Gartner

Why Your Storage Vendor Should Not Own Your Metadatabase

You do not buy your database from your storage vendor. There is no reason to buy your metadatabase from one either.

  1. Storage changes. Data does not. Enterprises replace storage every three to five years, but their data stays with them for decades. The metadatabase that describes that data has to outlive every array it sits on. A metadatabase sold with a storage platform is tied to that platform relationship.
  2. Every vendor catalog is another metadata silo. Each storage vendor now offers its own data intelligence layer. Run two or three storage vendors and you end up with two or three metadata catalogs, each with its own schema, APIs, and view of the truth. The organization trades storage silos for metadata silos.
  3. MCP across many catalogs means MCP bloat. When AI agents query several vendor catalogs through MCP, they have to learn each API, pull data from each, reconcile conflicting answers, and only then respond. That costs tokens, time, and accuracy. See MCP Bloat.
  4. Discovery is not management. Finding data on third-party storage is one step. Tiering it, migrating it, and moving it to AI without vendor-specific formats or a rehydration penalty is the part that saves money. Vendor-native tiering keeps data inside the vendor ecosystem, and moving it out means rehydrating it to primary storage first. Learn more about file-based tiering.
  5. Flash prices. An all-flash vendor has no incentive to tell customers that most of their flash holds cold data. Gartner estimates a 130% increase in combined DRAM and SSD prices by the end of 2026. The cheapest terabyte of flash is the one you do not buy. See Memflation. Learn more about how to stretch flash capacity.
  6. New versus proven. Everpure Data Intelligence was acquired in 2026 and its expanded AI capabilities became available in October 2026. Komprise has run in production with large enterprises for more than 10 years.

Storage vendors provide storage. Komprise provides unstructured data intelligence, data orchestration and data management, independent of any storage vendor. Buy any storage you want and keep full insight into your data and ensure no-lock-in data mobilization as your storage estate changes.

Tiering Signals vs Tiering

Everpure Data Intelligence surfaces tiering and reclaim signals. Blocks and Files reports those signals are passive: there is no optimized or policy-driven data placement.

Source: Blocks and Files

A signal tells you which data is cold. It does not move it off flash. Komprise data intelligence is built to act:

  • Find. Deep Analytics queries the Komprise Global Metadatabase across Everpure, NetApp, Dell, object and cloud storage by storage system, share, path, owner, file type, age, tag or KAPPA-enriched metadata.
  • Act. Save the query and it becomes the input to a data management plan that tiers files with Transparent Move Technology, confines them, or copies them, or to a Smart Data Workflow that scans for PII, enriches metadata or ingests curated data into AI.
  • Keep acting. The plan runs continuously as new data arrives. Tiered files move in native format with no rehydration penalty, users keep their original paths, and storage costs drop 70% or more.

Compression is not tiering either. Everpure says its new DeepReduce feature delivers a median 2:1 global data reduction on FlashBlade//E and FlashBlade//S200R2. Data reduction shrinks cold data, but the cold data still sits on flash and still sits in the backup set. Tiering removes it from both, and the two work together: reduce what stays on flash, tier what does not need to be there.

The Komprise Global Metadatabase: One Consistent Source of Truth Across Every Silo

AI needs one source of truth about enterprise data, no matter where that data resides: NAS, object, cloud, SaaS, or tape. If the AI sees only part of the data, or reconciles conflicting catalogs, it returns incomplete answers and hallucinates. The Komprise Global Metadatabase provides one consistent metadatabase across every data silo.

  1. Future-proof AI projects. Change storage vendors and your AI projects are unaffected, because the metadatabase does not live on the storage.
  2. Improve AI ROI. AI has access to all the data, regardless of where it resides, so results are complete rather than partial.
  3. Reduce MCP bloat. Agents use one set of APIs against one metadatabase, with no reconciliation across catalogs.
  4. End vendor lock-in. Change storage vendors at will. The Global Metadatabase stays intact.

Delivered as a Service, Proven at Scale

Here are some of the Komprise Intelligent Data Management advantages for unstructured data intelligence and data orchestration:

  1. No hardware, no kit, no GPUs. Some approaches hand you a kit to build out, with hardware to buy and staff to run it. Others require GPU-equipped hardware. The Komprise Global Metadatabase is delivered as a service. There is nothing for your team to build, deploy, or operate.
  2. Scales as data grows. The Global Metadatabase is a scale-out system. Komprise adds clusters as data grows, and each cluster is designed to ingest metadata from 10 billion files per day.
  3. Tested at enterprise scale. The Global Metadatabase works with hundreds of billions of files, and Komprise manages more than one exabyte of enterprise data across healthcare, life sciences, financial services, media, and public sector organizations.
  4. Heterogeneous by design. Komprise works across 15+ storage and cloud platforms, including Everpure, NetApp, Dell, Qumulo, VAST Data, HPE, IBM, Nutanix, AWS, Microsoft Azure, and Google Cloud.

Enrich With KAPPA, Query With MCP

A metadatabase is only as useful as the metadata in it. File system metadata covers size, dates, owner, and path. AI and analytics need more: project codes, patient IDs, instrument types, contract clauses, image attributes.

  1. Enrich with KAPPA data services. KAPPA data services is a serverless compute environment for extracting custom metadata across billions of files. Teams create a custom function, called a Kaplet, in minutes and run it only on the data they specify. KAPPA data services runs hundreds of Kaplets simultaneously and processes metadata for billions of files.
  2. Extend the metadatabase. Organizations define a custom extension of the core Global Metadatabase to store metadata specific to their business, populate it with Kaplets, and classify, curate, and orchestrate data based on that custom metadata to feed AI and analytics tools.
  3. Query with Komprise Universal File MCP. Ask questions in natural language about data on NAS, cloud, SaaS or tape from Claude, ChatGPT, Gemini or any MCP-compatible assistant, and act on the answer. See the next section.
  4. One connector, not a tangle. Dozens of storage-specific connectors and plugins are hard to maintain and hard to troubleshoot. Komprise provides one enterprise connector to all data, wherever it resides.

AI Agents on Everpure Data With Komprise Universal File MCP

Everpure MCP servers let agents query its data catalogs and its infrastructure topology and performance metrics. Komprise Universal File MCP, launched September 29, 2026, answers a different question: not how the arrays are performing, but what is in the data and what to do with it, across every storage vendor.

  • One interface, one schema. Agents query the Komprise Global Metadatabase instead of loading a separate MCP server and tool set for each storage system and catalog, which is the MCP bloat that raises token costs and lowers accuracy.
  • Right-sized answers. Progressive response returns metadata first through Transparent File Tables and loads full files only when the task needs them. Deep Analytics filters out irrelevant, outdated, conflicting and duplicate files, and filtering out ROT data improves AI accuracy by over 50%.
  • Governed by default. Every request authenticates through existing SSO and returns only data the requesting user is already permitted to see. Komprise audits what data was sent to which AI tool.
  • Answers that turn into action. From any result, a user or agent triggers a Smart Data Workflow: confine sensitive files, run a KAPPA data service to enrich metadata, ingest a curated set into an AI pipeline, or export an Apache Iceberg table for Databricks or Snowflake.
  • Tiered data included. Files tiered off FlashBlade or FlashArray with Transparent Move Technology stay queryable with no rehydration.

Source: Komprise Universal File MCP

Source: SiliconANGLE

KAPPA data services enriches. Universal File MCP queries, learns, and orchestrates.

From Metadatabase to Databricks and Snowflake

The Komprise Global Metadatabase connects directly to Databricks and Snowflake. Run a query through Universal File MCP or in Komprise to curate a subset of the metadatabase, and Komprise creates an Apache Iceberg table with pointers to the files behind each record using Transparent File Tables. Load it into Databricks or Snowflake and join ERP and other business tables with file metadata, analysis that was not possible when file metadata was locked in storage silos.

How Komprise Works With Everpure Storage

Komprise treats Everpure file and object storage as one of many sources and targets.

  1. Analyze. Komprise indexes file metadata on FlashBlade and other NAS and object systems into the Global Metadatabase and shows data age, access, owners, and growth across all of them.
  2. Assess before buying. The Komprise Flash Stretch Assessment identifies how much primary flash capacity can be reclaimed, up to 70%, which is worth more than $350,000 per petabyte at current prices.
  3. Intelligent Tiering. Transparent Move Technology tiers cold files from Everpure flash to lower-cost object or cloud storage at the file level, with no rehydration penalty. TMT cuts 70%+ of storage costs. FlashBlade also serves as an S3 target for tiered data.
  4. Migrate to, from, or between Everpure systems. Elastic Data Migration moves file and object data onto FlashBlade or FlashArray from other NAS platforms, between Everpure systems during an Evergreen upgrade or consolidation, or from Everpure to the cloud. It delivers 27X faster NFS migrations, and Komprise Hypertransfer delivers 25X faster SMB migrations over the WAN. Migrations run in parallel with checksums for data integrity, and permissions and metadata move with the data. Tier cold data first, and only the active data needs to move onto new flash.
  5. Prepare data for AI. Deep Analytics and Smart Data Workflows find, classify, and curate unstructured data across Everpure and non-Everpure storage, and Komprise Intelligent AI Ingest delivers curated, governed data sets to any AI stack with 2X faster data flows.

Everpure Data Intelligence vs the Komprise Global Metadatabase

Evaluation Criteria Everpure Data Intelligence Komprise
Who provides the metadatabase Your storage vendor A data management company independent of any storage vendor
Maturity 1touch acquired in 2026; expanded release available October 2026 In production with large enterprises for more than 10 years; more than one exabyte managed
Delivery model Part of the storage vendor platform and relationship Delivered as a service; no hardware, no kit, no GPUs to buy or run
Scale Not publicly specified Hundreds of billions of files; each scale-out cluster designed to ingest metadata from 10 billion files per day
Discovery coverage Everpure arrays, other on-premises storage including mainframes, public cloud and SaaS NAS, object, cloud and tape across 15+ platforms in one consistent Global Metadatabase
Multi-vendor reality One more vendor catalog alongside each other vendor catalog One metadatabase across vendor silos
Metadata enrichment Classification and semantic context KAPPA data services runs hundreds of custom Kaplets across billions of files; custom metadatabase extensions
Tiering signals vs action Passive tiering and reclaim signals; no policy-driven data placement Saved Deep Analytics queries become continuous tiering, confinement and copy policies on any storage
Data reduction vs tiering DeepReduce compresses data that stays on flash Cold files leave flash and the backup set entirely
AI agent access Native MCP to Data Intelligence catalogs; MCP servers for Fusion and Pure1 infrastructure metrics Komprise Universal File MCP: one SSO-governed, audited MCP to data across all storage, with workflow triggers and access to tiered files
Progressive response for AI Catalog queries Metadata first through Transparent File Tables; full files only when needed; ROT filtered out
Analytics integration Not specified Transparent File Tables create Iceberg tables for Databricks and Snowflake
Data migration Focused on the Everpure platform Elastic Data Migration, 27X faster NFS; Komprise Hypertransfer, 25X faster SMB; onto, between, or off Everpure
Rehydration penalty Applies when tiered data leaves the vendor ecosystem None for access or migration
Flash price exposure Growth met with more Everpure flash as NAND prices rise 15% to 20% in 4Q26 Flash Stretch assessment reclaims up to 70% of primary flash first; 70%+ savings with Transparent Move Technology
When storage changes Metadata and policies tied to the vendor relationship Global Metadatabase stays intact; AI projects unaffected

Everpure and Komprise FAQ

What is Everpure?

Everpure is the new name of Pure Storage, effective February 23, 2026. The company makes FlashArray and FlashBlade all-flash storage and now positions itself as a data management company with its Enterprise Data Cloud platform and Everpure Data Intelligence, formerly 1touch.

Is Pure Storage now called Everpure?

Yes. Pure Storage renamed itself Everpure on February 23, 2026. The stock ticker stayed PSTG, and existing products, contracts, and support terms did not change. Product names are moving to Everpure branding through 2026.

Why did Pure Storage change its name to Everpure?

The company renamed itself to signal a shift from storage hardware to data management. Everpure positions its platform as an Enterprise Data Cloud and added data discovery and classification by acquiring 1touch.

What is the Everpure Enterprise Data Cloud?

The Enterprise Data Cloud is the Everpure architecture for managing storage and data across on-premises and public cloud environments as a service, with a centralized control plane over Everpure arrays.

What is Everpure Data Intelligence?

Everpure Data Intelligence is the data discovery and classification software Everpure gained by acquiring 1touch in 2026. Available in October 2026, it discovers and classifies data across Everpure arrays, other on-premises storage, public clouds and SaaS applications, flags PII and PHI, and offers native MCP access to its catalogs. Its tiering and reclaim signals are passive, with no policy-driven data placement.

Why does AI need one consistent metadatabase?

AI needs one source of truth about enterprise data. With several vendor catalogs, AI sees partial data or must reconcile conflicting answers, which lowers accuracy and drives hallucinations. One consistent metadatabase across NAS, cloud, SaaS, and tape gives AI access to all the data through one set of APIs.

Does storage vendor data management create vendor lock-in?

Yes. Data intelligence and data tiering sold with a storage platform keep metadata, policies, and tiered data within that vendor relationship, so changing vendors means rebuilding the analysis and rehydrating tiered data back to primary storage. Storage-independent software such as Komprise keeps the metadatabase and policies portable and tiers whole files with no rehydration penalty.

Does Everpure Data Intelligence tier data?

No. Blocks and Files reports that Everpure Data Intelligence provides passive tiering and reclaim signals, with no optimized or policy-driven data placement. Acting on those signals still takes a separate tool. Komprise turns a saved Deep Analytics query into a continuous tiering policy with Transparent Move Technology, across Everpure and every other storage system.

Does Everpure Data Intelligence tier data?

No. Blocks and Files reports that Everpure Data Intelligence provides passive tiering and reclaim signals, with no optimized or policy-driven data placement. Acting on those signals still takes a separate tool. Komprise turns a saved Deep Analytics query into a continuous tiering policy with Transparent Move Technology, across Everpure and every other storage system.

Should Everpure customers buy more flash or tier cold data first?

Tier first. About 70% of unstructured data on primary storage is typically cold with a high percentage of ROT data, and enterprise SSD prices hit record quarterly increases in 2026. The Komprise Flash Stretch Assessment identifies how to reclaim up to 70% of primary flash capacity, worth more than $350,000 per petabyte at current prices, before an expansion or upgrade.

How do you reduce FlashBlade or FlashArray storage costs?

Identify cold files with storage-independent analytics, then tier them at the file level to lower-cost object or cloud storage while users keep access from the original location. About 70% of unstructured data on primary storage is typically cold, so tiering before a storage refresh, an expansion or Evergreen upgrade reduces the flash capacity you pay for.

How do you migrate data to Everpure FlashBlade or FlashArray?

Analyze the source data first, tier cold files to lower-cost storage, then migrate only active data. Komprise Elastic Data Migration moves NFS data 27X faster and Komprise Hypertransfer moves SMB data 25X faster, from any NAS vendor or cloud onto Everpure, between Everpure systems, or from Everpure to the cloud. Because Komprise is storage-independent, the same Global Metadatabase and policies continue after the migration.

How does Komprise work with Everpure storage?

Komprise indexes file data on Everpure storage alongside other NAS, object, cloud, SaaS, and tape sources in the Global Metadatabase, tiers cold files off flash with Transparent Move Technology, migrates data onto, between, and off Everpure systems with Elastic Data Migration, uses FlashBlade as an S3 target, prepares data across all storage for AI, and gives AI agents one MCP interface through Komprise Universal File MCP. Customers reclaim up to 70% of primary flash capacity.

How is Komprise different from Everpure?

Everpure sells all-flash storage and a data intelligence layer tied to that storage relationship. Komprise is a storage-independent data intelligence and orchestration layer powered by the Global Metadatabase. Everpure Data Intelligence surfaces tiering signals; Komprise acts on them, tiering cold files off flash in native format with no rehydration penalty and running continuous policies across every vendor. Komprise has run in production with large enterprises for more than 10 years.

Why do I need Komprise if I have Everpure?

An all-flash vendor stores your data on flash. Komprise makes sure only the data that needs flash stays there. You get a Flash Stretch assessment that reclaims up to 70% of primary flash before an expansion, file-level tiering that cuts storage costs 70% or more, one metadatabase across vendors instead of another vendor catalog, and one governed MCP interface for AI agents across all storage.

How do I migrate data off Everpure FlashBlade or FlashArray?

Komprise Elastic Data Migration moves file and object data from Everpure to another vendor or the cloud, 27X faster for NFS, and Komprise Hypertransfer moves SMB data 25X faster, with permissions, metadata and checksums preserved. Analyze first and tier cold data before cutover, so only active data moves. The Global Metadatabase and policies carry over unchanged.

How do AI agents access Everpure data with Komprise?

Komprise Universal File MCP gives AI agents one MCP interface to Everpure and every other storage system through the Global Metadatabase. Answers are scoped to the user’s existing permissions, start with metadata and load full files only when needed, and exclude irrelevant, outdated and duplicate files. Agents can trigger Smart Data Workflows from results and reach tiered files with no rehydration.

Want To Learn More?

Related Terms

Getting Started with Komprise: