NetApp and NVIDIA are teaming up to help teams actually run, trust, and scale enterprise AI in production. In the second episode of the AI Intelligence Lab video series, Arindham Banerjee, Chief Platform and Technology Officer, speaks with Kari Briski, VP of Generative AI Software for Enterprise at NVIDIA, about the importance of contextualized, enterprise-specific data over generic internet data for relevant, trustworthy AI outcomes.
What the AI Data Engine Actually Does
NetApp AI Data Engine combines semantic search, data vectorization, and data services into a Retrieval-Augmented Generation (RAG) pipeline that runs on governed, production data. By integrating NVIDIA NIMS, NeMo, and the NVIDIA AI Data Platform reference architecture, the solution grounds AI responses in an organization’s own version-controlled documents, tables, and rich PDFs rather than relying solely on public internet content. This provides them with more accurate, trustworthy outcomes that are tied to the latest state of enterprise data, not stale snapshots.
From AI Pilots to Real Production
In the episode, Arindham and Kari spend most of their time discussing how to help customers finally move from endless pilots to live, production AI. Many enterprises experiment with open and proprietary models and basic RAG setups but get stuck on issues such as hallucinations, incorrect document versions, or outdated data. The AI Data Engine relies on a validated, end-to-end reference design and built-in governance and data services from NetApp and NVIDIA, so teams can move forward with confidence rather than rebuild the stack themselves. Users query with natural language (including “talk to your data”–style experiences) while the platform handles all the complexity of reasoning over structured data (writing and fixing SQL) and unstructured content in the background.
Governance, Security, and Trusted Outcomes
Enterprise AI only works if security and compliance teams can sleep at night. The AI Data Engine puts governance to the forefront, using roughly 180 classifiers to enforce privacy and policy rules so data officers know sensitive information is protected. Every answer is grounded in governed data, with clear controls over who can access information and how it’s used, reducing the risk of models leaking or misusing corporate information. By pairing real-time access to production data with built-in guardrails, the AI Data Engine gives organizations the confidence to trust AI in everyday operations.
Turning Storage Into Intelligence
Arindham and Kari explain how storage is no longer just about holding bits and bytes. Every stored object becomes a source of intelligence. As AII generates insights, those insights become assets in their own right and turn storage from a cost center into a driver of business value. Because moving petabytes or exabytes of unstructured data isn’t practical, the focus shifts to “bringing AI to the data.” By enabling near-data compute and transforming information in place at internet scale, organizations can unlock value without massive data movement. Over time, co-innovation between NetApp and NVIDIA will continue to refine these architectures for faster retrieval, smarter generation, and continuous learning from how users interact with their data.
The Road Ahead: Data Intelligence at Internet Scale
Looking forward, NetApp will evolve from the number one storage provider to the number one data intelligence provider, and the AI Data Engine is the starting point. The vision is to make enterprise data searchable, governable, and discoverable at scale across on-prem, cloud, and everything in between by transforming every byte into something that can be indexed and reasoned over. That includes building distributed knowledge graphs and a single indexable view across all data types and modalities, from text to video to audio. With more than 30,000 customers sitting on massive data estates, both companies see huge upside in helping them “talk to their data,” move from pilot to production, and spin up a data flywheel where every interaction feeds better, more personalized AI over time.
Explore More of the AI Intelligence Lab
In Episode 3, the conversation continues about intelligent data services, why AI creates new data management challenges, and how storing semantic understanding of data is the new attack surface for threat actors.