Hitachi Vantara positions intelligent storage as the missing link in AI production
Companies scaling artificial intelligence from experimental phases to operational deployments are discovering that managing data—not computing power—presents the steepest challenge, with open table formats like Apache Iceberg emerging as a key solution.

As enterprises move artificial intelligence initiatives beyond proof-of-concept stages into full production, they face a fundamental obstacle: data management, not GPU capacity, represents the real bottleneck. Open table formats including Apache Iceberg are helping address this challenge. While demand for AI infrastructure continues to accelerate, the information assets companies require remain fragmented across multiple systems, lacking proper governance and accessibility.
This reality is reshaping how organizations think about storage infrastructure. Storage systems are evolving into intelligent layers capable of categorizing, managing and delivering data directly to machine learning models. Unstructured content—including system logs and medical imaging files—contains valuable signals that modern AI can now extract, according to Arun Prasad Raman, vice president of hybrid cloud and AI infrastructure at Hitachi Vantara Corp.
Storage is not seen as just storing the data. You need to bring in more intelligence natively rather than moving the data to another replication and bringing it back for, say, indexing. If it can do it natively, that's where the value add is.
Arun Prasad Raman, Hitachi Vantara
Raman and Jeff Lundberg, principal product and solutions manager at Hitachi Vantara, shared these perspectives during an exclusive conversation at SC26 with theCUBE Research's John Furrier and Gemma Allen, broadcast on theCUBE, SiliconANGLE Media's livestreaming platform. Their discussion centered on bridging the gap between AI pilots and production deployments, the role of Apache Iceberg and Hitachi Vantara's approach to embedding intelligence within storage systems.
Apache Iceberg and the shift toward intelligent storage
Reference designs for AI infrastructure and data platforms have largely converged on common patterns. Yet many initiatives encounter obstacles because data resides across disparate environments—on-premises infrastructure, cloud services and third-party software platforms—while compliance requirements, access controls and governance protocols create additional complexity, according to Lundberg. Hitachi Vantara is addressing these challenges through purpose-built solutions leveraging its Hitachi iQ stack, supplemented by domain expertise from affiliated businesses in rail and energy sectors.
It's about the outcome more so than the technology that we're trying to put into place. That's an enabler to get you to the solution you're trying to get to.
Jeff Lundberg, Hitachi Vantara
Much of this transformation centers on storage capabilities. Hitachi Vantara recently enhanced its VSP One platform with new object storage and data lakehouse features. Within its object storage service, a capability automatically identifies files containing sensitive information such as Social Security numbers. Data scientists gain the ability to exclude that information from particular models or prevent its transfer across geographic boundaries, Raman explained.
That's the intelligence we are adding into the storage itself. So that's where the evolution of storage is making it more intelligent.
Arun Prasad Raman, Hitachi Vantara
The same principle applies to analytical workloads. Medical imaging metadata—such as CT scan information—can be stored using Apache Iceberg format on the company's object storage and accessed through direct queries. A healthcare application could retrieve all abdominal scans from the preceding twelve months without requiring a separate extract, transform and load operation or intermediate database layer, Raman noted.
https://www.youtube.com/embed/D4_V-p0RkVc?feature=oembed
Whether it's a backup or it's 10 years old, it still carries some intelligence. As long as it carries intelligence, it becomes important to AI.
Arun Prasad Raman, Hitachi Vantara