London, September 30, 2026 / PRNewswire / -- Everpure ( NYSE : P) Today, we announce the launch of new platform capabilities aimed at simplifying enterprise data management and advancing our "Data First" vision. This concept was introduced in June of this year at Pure // Accelerate. Its core idea is that in the era of AI, the driving force behind enterprise architecture must be data, not applications. The company states that today's update will continue to enhance modern data management capabilities in order to address the bottlenecks that hinder the scaled implementation of enterprises, including fragmented contexts, unpredictable inference costs, and slow and complex deployments.

"Enterprise AI is facing bottlenecks, not because the models are not strong enough, but because the data is not yet ready for real-time, autonomous proxying," said Everpure, General Manager of Data and Digital Experience at Prakash Darji. "We are eliminating this friction. By ensuring that enterprise data is continuously governed, automated, and readily accessible, we provide the foundation and confidence that organizations need to move AI from the laboratory to the production environment."

Make enterprise data management simpler and more secure.

Enterprises that deploy AI proxies often find it difficult to grant them reliable and governed data access across their entire operations, which limits the capabilities of these proxies. Everpure Data Intelligence addresses this issue by detecting, categorizing, and establishing enterprise information context at the source, covering Everpure platforms, public clouds, SaaS applications, as well as third-party storage. Based on this, new capabilities enable autonomous proxies and administrators to directly and securely access real-time enterprise context without the need to customize API processes:

• Native MCP integration:Implement an open model context protocol (MCP), allowing AI proxies and security tools to use natural language to query real-time data directories. Proxies can now search for relevant data and understand its sensitivity level, which serves as input for AI, proxy workflows, and analysis.
• Plug and play deployment:Simplify deployment through the existing Pure1 console; there is no need for complex professional service projects, which reduces the operational overhead associated with setting up independent management servers and accelerates the time to value realization.
• Privacy-first file intelligence:Without reading the contents of the files, it displays who has access to each file share and whether the data is outdated, thereby helping the team to fix any exposed risks and reclaim capacity before opening the shares to the AI proxy.

Improve the performance of AI, next-generation efficiency, and cost predictability.

To support the growing workload from archiving to AI, Everpure is expanding the boundaries of performance and capacity efficiency in a non-destructive manner. New capabilities bring high-performance AI execution directly to the source data, providing production-level speeds without migrating the data from its original location in the system:

• Accelerate LLM reasoning through PureKVA ( Key - Value Accelerator ):Everpure FlashBlade Now, it is possible to directly pre-set the context into the GPU memory, increasing the initial token output time ( TTFT ) by up to 20 times. This solution supports multi-tenancy for enterprises without the need to migrate datasets, which can eliminate GPU idle time, increase token throughput, and reduce response latency in real-time applications.
• Always online DeepReduce data compression:On the FlashBlade system, continuous scanning of storage blocks is performed to identify sub-block-level data similarities that traditional deduplication methods cannot detect, even for compressed content. The available storage capacity will automatically expand without affecting write performance, and there is no need for manual scheduling. This significantly reduces hardware usage and cross-cloud costs.
• Intelligent token Optimization Reference Architecture:In light of the need for enterprises to take control of their own data and optimize AI expenditures, Everpure now provides a reference architecture that utilizes an open-weighting model. This gives enterprises more control over their data, makes AI costs more predictable, and at the same time reduces the overall usage of external providers API and token.

The company stated that these capabilities, considered individually, address specific gaps in simplification, security, performance, and cost; when combined, they provide a unified and continuously updated foundation for enterprises to manage data and run AI in a production environment, and they can evolve in tandem as the scale of proxy-based workflows expands. Since these capabilities identify which data is sensitive and who is accessing it, they also enhance network resilience. Ensuring that data can be securely used within the same context of AI, they also determine how the data should be protected and which content should be prioritized for recovery in case of an outage. The company announced that the new capabilities will be launched in October of this year.

More resources

• Executive View: From Data Processing to Data Priority: The Proxy Inversion of Work and the IT Architecture
• Press release: Everpure releases a data-first architecture for the AI era in June 2026
• Blog: Your AI strategy is blocked by the application center architecture.

About Everpure

Everpure ( NYSE : P) helps organizations take control of their data through industry-leading, continuously evolving storage and data management platforms. The company assists clients in unlocking the value of their data, ensuring its accessibility and intelligence, and preparing them for the AI era. Everpure makes data management easier while enhancing performance and significantly reducing energy consumption. The company claims that its Net Promoter Score ( NPS ) has remained high for over a decade, and it is the choice of the world's most innovative organizations. For more information, please visit www.everpuredata.com.

This press release contains forward-looking statements regarding the business and product roadmap of Everpure. The timing, development, and release of any Everpure products, features, or characteristics mentioned in this text are solely at the discretion of Everpure. The information provided is for reference only and does not constitute a commitment, guarantee, or legal obligation to deliver any materials, code, or features. The content discussed should not be used in purchasing decisions, nor should it be incorporated into any contracts. The performance indicators mentioned are for reference only and are not performance guarantees; actual results may vary significantly from those described in this document due to differences in deployment environments and dataset conditions.