Redefining Data Infrastructure for AI with NetApp

At the Tech Field Day Experience during NetApp INSIGHT 2025, Syam Nair, Chief Product Officer at NetApp, outlines their strategy to redefine data infrastructure for artificial intelligence (AI). The focus is on building intelligent storage systems that unify data across block, file, and object formats, enabling fast and secure access to AI-ready data. By integrating AI capabilities directly into the storage layer, such as metadata enrichment, tokenization, and data governance, NetApp aims to empower storage administrators and extend the usability of data for AI workflows without sacrificing control or security.

In his presentation, Syam Nair emphasized that the AI data landscape is shifting from hype to practical application, with growing unstructured data sources such as machine-generated and generative AI outputs. With the ONTAP platform at its core, NetApp’s vision hinges on making all data AI-ready by embedding intelligence—like security policies, tokenization, and embedding mechanisms—into the data layer itself. This not only ensures data accessibility and governance but also removes the need for complex extraction, transformation, and loading (ETL) processes. NetApp’s AI Data Engine (AIDE) and AFX are designed to streamline this intelligent access while reducing the proliferation of data copies by managing metadata and vectorization in place.

NetApp’s approach aims to elevate the role of storage administrators, transforming them from infrastructure caretakers into enablers of data-centric applications and AI workflows. Instead of pushing AI users to understand the storage backend, NetApp provides APIs and policy-driven data access mechanisms that integrate with tools like Kafka or database systems. Emphasis was placed on security through granular, zero-trust policies and governance over metadata to prevent overhead and sprawl. NetApp aims to support emerging standards such as Apache Iceberg for semantic access and to evolve toward a system where unstructured data can be consumed like structured data—offering semantic reads without altering write formats. Ultimately, NetApp is not attempting to replace databases but rather to unify and enrich data access directly within the intelligent storage infrastructure.