Dell’s president of ISG on how the company cashed in on the AI boom

If Dell doesn’t immediately come to mind when thinking of major AI players, nobody would blame you. The company has moved like the g in lasagna away from laptops and other consumer goods and into AI infrastructure.

A core component of this push is Dell’s enterprise-focused Infrastructure Solutions Group (ISG) which has used a familiar Dell playbook to grab pole position in the AI race. It is creating a comprehensive AI needs ecosystem for customers who want end-to-end solutions (think servers, storage, and networking solutions).

Revenue Brew spoke to Dell’s Arthur Lewis, president of the ISG unit, about the pivot, and what comes next.

This conversation has been edited for length and clarity.

When the pivot towards AI started, what was the original strategy? How did it evolve over time?

We understood there was going to be a lot of training infrastructure that needed to get built around these foundational models, and the way in which we typically go to market, the way that we typically approach engineering and product roadmaps, was really not going to carry the day.

We have uptimes of over 99% and when you get into this complicated infrastructure, that really says a lot. Our ability to design, our ability to deploy and integrate our quality, our service and support, you couple with that our financing capabilities that we have for some of these very large orders. You start to stitch together a really nice story for customers, and that’s part of the reason and why we stay very advanced.

What was Dell already doing that allowed it to have the skills to make this pivot to AI?

We’re in our 42nd year of existence, Michael [Dell] has been very big in driving, listening to the customers. We have direct relationships with many customers. We’re talking to them on a daily basis.

AI is a different problem statement for a lot of customers, and they don’t necessarily have the skill they need to think about near-term, mid-term, long-term strategies in order to deploy AI effectively.

We’re well poised, not only with the neoclouds, but also with what is like the biggest opportunity over the next decade: the enterprise and the inference workload.

I’m glad you brought up enterprise because it seems clear Dell has a good reputation in that space. How do you keep that reputation going as you move into enterprise AI?

We have transitioned from being a fast follower to being an innovation leader in AI. Whether you think about the neoclouds, whether you think about sovereigns, whether you think about the enterprise, we have become very opinionated and forward-leaning into the architecture of the future, which I think customers not only need, but require, because they don’t always have all of the answers.

When you think about your partnership approach, how do you make sure Dell remains an essential orchestrator?

We stay very close, not just with what’s in the market today, but what’s coming next month, next quarter, second half, next year—really making sure that we provide a broad swath of optionality for customers.

Given how fast AI forces your team to change, how do you make sure you’re continuing to stay with that pace while also providing the service that Dell is very much known for?

We have relationships with all the large foundational models. We meet very frequently, to understand what they’re doing, what they’re seeing. We have advanced versions of their software so we can play with it and test it. We stay very well connected across the industry, including AI startups.

We knew that artificial intelligence was going to move quickly. But think about the fact that we’ve gone—in 36 months—from these one-shot inference models to these long-thinking autoregressive models, to this more sophisticated agentic technology.

What excites you the most about how you could build on the early success in the AI category?

Artificial intelligence is going to require a data center that looks a lot different in the future than what it looks like today. There’s a strong proclivity for customers to have access to all of their data, and to do it where their data resides, which is on-prem. That creates a massive infrastructure opportunity over the next couple of decades.

What would you say is the most underrated part of what Dell has done to get to where it is in the AI race?

I would say probably our storage portfolio. When you take a look at what we’ve done in private cloud, when you take a look at what we’ve done in our unstructured world, what we’ve done with our AI data platform, and the acquisition of Dataloop and what we’ve done there—that’s probably maybe less understood, but we’re working hard to make sure people understand the full stack solution.