NetApp INSIGHT 2025: The Enterprise Data Platform—Intelligent, Resilient, Disaggregated | Gagan Gulati, NetApp

Clips
More from NetApp INSIGHT 2025: The Enterprise Data Platform—Intelligent, Resilient, Disaggregated

Gagan Gulati

Senior VP and General Manager for Data Service

NetApp

play_circle_outline Introduction of Gagan Gulati, GM of Data Services at NetApp, at NetApp Insights 2025.

play_circle_outline Enhancing AI Performance: The Critical Role of AFX and NetApp's Data Platform with NVIDIA Integration

Info
Transcript

Gagan Gulati, NetApp

Gagan Gulati

Senior VP and General Manager for Data Service NetApp

At NetApp INSIGHT 2025 in Las Vegas, theCUBE Research’s Christophe Bertrand sits down with Gagan Gulati, general manager of data services at NetApp, to unpack how NetApp is evolving into an enterprise data platform for AI. Gulati details the new AFX system – true disaggregated storage that independently scales capacity and performance – and how pairing it with the AI Data Engine (powered by NVIDIA) enables an end-to-end AI pipeline that data scientists and engineers can operationalize with confidence.

The conversation explores AI as the “first true h... Read more

explore Keep Exploring

What is your perspective on the role of AFX in NetApp's evolution as a data platform, especially in relation to the recent innovations around AI? add

What recent announcements were made by NetApp regarding their enterprise data platform for AI? add

bolt Powered by CUBE AI

Gagan Gulati, NetApp

search

Christophe Bertrand

>> Welcome back to Las Vegas, NetApp Insights 2025. I'm Christophe Bertrand, Principal Analyst with theCUBE Research. I'm very, very pleased to have you here, Gagan Gulati. You are the GM of Data Services, right? I got that right. We meet again. We met a few months ago at Analyst Event. Had a great conversation. Obviously, today I want to talk about the announcements on the AFX. Clearly, lots of innovations around AI. The AFX is a very critical piece of the equation. George is talking now about NetApp really becoming a data platform. AFX has to be at the heart of that. So tell us more. What's your view, what's your perspective, and how did we get to this point?

Gagan Gulati

>> Yeah, look, as George mentioned, it's all about data and NetApp is the only enterprise data platform for AI. And what we announced today is one, of course, our AFX system, which is the true disaggregated storage that allows you to independently scale your capacity and performance. And I think it's a game changer. What we also therefore announced today is also the ability to build an end-to-end AI pipeline using the power of AI Data Engine. And when you put both of them together, you get a very solid portfolio that our customers have been telling us that they love from us, not only access programs as they can put AFX together now with AI Data Engine that's powered by NVIDIA to get the best of what they want for their data scientists and their engineers. So we are extremely happy and extremely proud and it's been a long time coming for us.

Christophe Bertrand

>> So let's switch gears a little bit and look at the other component of the conversation, which is it's not just AFX, it's really a hybrid world out there. And there were also a number of announcements and conversations around cloud providers, hyperscalers, and you've been in partnership with most of them for many, many years to the point where you could almost claim there's a level of cloud neutrality or hyperscaler neutrality in a sense. You can get the same feature functions across the cloud of your choice. Well, in the context of AI and the evolution that we are seeing, we're about to see more, how does that play out? How do you think, from your standpoint, those investments are going to pay out and why is it so critical that you have these relationships?

Gagan Gulati

>> Yeah. Look, AI is the first true hybrid worker, right? And that by default means that you have to look at AI a little differently. A lot of the projects today, they will start in cloud, especially when you're doing pilots. And data gravity, which is where your actual sensitive data is, the data that's actually going to be used at influencing time at run time, that pretty much will define where your AI is going to be. You're going to basically bring AI to data. You're not going to move your data to AI. And when that happens, you want to make sure that your data infrastructure and your data platform is able to support the needs of what our customers therefore want, which is whether your data is in your own data centers or your data is in any of the hyperscaler clouds or AWS as your MTCP and GCP. And our partnership really helps there because we are natively built in all three of them helping our customers out.

Christophe Bertrand

>> So I totally get that. It makes perfect sense. We'll see how things evolve because I have a suspicion compliance and governance may affect the equation a little bit in some geos. We'll see about that. Cyber resilience, I know one of the topics we like to talk about, and I like to research is probably another driver here, but I totally agree with the perspective. One of the things, though, that I read about, and I've mentioned this a few times here on theCUBE today is a lot of projects don't really get to ROI. Actually, I think MIT did research recently and said, "Well, 95% or more of projects don't really generate an ROI," which, well, makes you wonder, there's so much investment, so much hype, so much attention. Why do you think that's the case? And then how are you going to help unlock the ROI? I have a suspicion I know, but I'm curious about why you think that's happening today.

Gagan Gulati

>> Yeah. Look, I think you're absolutely right and the studies basically show that a vast majority of these projects do not make it to production, and they don't make it to production because AI is data-hungry by default. To get to the right AI model, you need to give it the right data. And if the data itself and the data infrastructure itself is not ready, your AI is not going to be ready. That's a fundamental problem. And when we met with hundreds of our customers to understand really what's going on, they explained to us the four big issues. One around is, of course, data visibility. For data scientists to create the right data sets that can be used for AI, you first need to know where your data is. And most of our customers, we know that over decades have data sprawled everywhere. Could be in on-premises, data centers, could be everywhere. And as an organization, we have no idea where that data is specifically that your data scientists and data analysts want. So that's the first big problem, which is actually visibility of data. You need to be able to get that visibility and then you also have to make sure that this data is well protected, that's always available, and that it's not that complex to get to it. That's number two. Then you have the typical problem of security and governance. Before you make this data available for AI, you want to make sure that it's sensitized. You don't want to have highly sensitive customer data going into the pipeline. If you're running an AI project, let's assume against patient EMRs or electronic medical records, you want to make sure that the patient IDs and social security numbers don't make it to the pipeline. And last but not the least, all about the cost. You want to make sure that these AI projects are not that expensive and the data bloat stays. So I think that fundamentally, most of these projects do not graduate from pilots to actual production because of these four issues. And this is where NetApp excels in because we have hundreds of exabytes of data, 30 years of innovation, and what we announced today with AFX and AIDE helps our customers with the entire journey.

Christophe Bertrand

>> And you'll be applying it to yourself as well as part of the process. I'd like to talk about cyber resilience before we close, and I know you've done it quite a bit around cyber resiliency following the NIST model. Certainly have researched the space. It's terrible. A lot of help is needed. And that's going to be a key element, to your point. Data has to be resilient, has to be compliant, et cetera. In a nutshell, tell us about how you're helping customers and what you're looking forward to doing moving forward in terms of maybe innovations and how AI may play into that as well.

Gagan Gulati

>> Yeah. Look, the word AI has actually made the whole ransomware problem worse because the attackers are now using AI, they're most sophisticated than ever before. The attacks are getting worse. And this is the world of double extortion attacks, which is just not about encrypting data and then asking ransom for the key to decrypt it, but it's also about making a copy of the data and run away and then asking for more ransom to give you the data back, which you may never get back. So what we are doing at NetApp is to help our customers with that overall picture of ransomware resilience. And today we announced two big features. One we call as the data breach detection capability. We are the first in the industry to do that, very, very proud of it. And number two is isolated recovery environments because you want to make sure that not only able to detect the attacks, you want to recover that with confidence so that you don't get reattacked if those backup copies or snapshots have malware. So we're bringing a lot of AI innovations into our ransomware resilience suite of products, and I hope that our customers get to see all of that.

Christophe Bertrand

>> And that's what makes you the most secure storage on the planet with guaranteed recovery. You should make a t-shirt. You should make a t-shirt. Gagan, thank you so much.

Gagan Gulati

>> Thank you for that.

Christophe Bertrand

>> Thank you so much for joining us. It was a pleasure having you here on theCUBE.

Gagan Gulati

>> Thank you.

Christophe Bertrand

>> And to our viewers, stay tuned. NetApp Insights live from Vegas. I'm Christophe Bertrand, Principal Analyst with theCUBE Research. We'll be right back.