Dell Technologies World 2026 | Bob Ward, Microsoft & Robert Sonders, Dell Technologies

Clips
More from Dell Technologies World 2026

Bob Ward

Principal Architect

Microsoft

Rob Sonders

Technical Staff Engineering Technologist

Dell Technologies

play_circle_outline Dell Launches Automation Studio and SQL Server 2025 with In-Engine AI, Integrating Azure Local, Dell Private Cloud, NVIDIA, On-Prem

play_circle_outline Private AI Deployments in Finance, Manufacturing and Healthcare: Enabling Secure Edge Clinics and Local Clinical Workflows

play_circle_outline Sovereign Hybrid Management with Azure Arc: Automated Day‑One/Two Firmware and Hardware Lifecycle Across Sites

play_circle_outline SQL Server as secure data plane enabling AI on-prem without moving data.

Info
Transcript

Bob Ward, Microsoft & Robert Sonders, Dell Technologies

Bob Ward

Principal Architect Microsoft

Rob Sonders

Technical Staff Engineering Technologist Dell Technologies

In this interview from Dell Technologies World 2026, Bob Ward, principal architect at Microsoft, joins Robert Sonders, technical staff global engineering technologist, platform automation and multicloud software, at Dell Technologies, to talk with theCUBE's John Furrier and theCUBE + NYSE Wired's Gemma Allen about how enterprises are moving AI from experimentation into full-scale production through hybrid infrastructure and a unified data platform. Ward highlights the general availability of SQL Server 2025, now featuring native AI capabilities — including ve... Read more

explore Keep Exploring

What are the AI features and deployment options in SQL Server 2025 (SQL AI), and how does it integrate with Dell platforms and on‑premises model runtimes? add

In which industries or environments are you seeing these opportunities play out? add

How do Dell and Microsoft integrate Dell's automation platform, SQL Server, Dell Distributed Private Cloud (including Azure Local) and Azure Arc to support AI agents/an AI operating model for edge deployments (e.g., hospitals), and what factors drive customers to choose Azure public, Azure Local, or a self‑managed Dell private cloud? add

What are your thoughts on the data challenges around SQL, and how does embedding AI in the database change the need to move SQL data for AI processing? add

bolt Powered by CUBE AI

Bob Ward, Microsoft & Robert Sonders, Dell Technologies

search

Gemma Allen

>> Welcome back to The Cube. We are coming to you here from Las Vegas. We are covering Dell Technology World 2026 and we are going under the hood of what exactly is happening as AI moves from experimentation to deployment. I'm joined by my co-host, John Furrier and joining me now are two folks who have been in this partnership for quite a while. Bob Ward, principal architect at Microsoft and Robert Sonders, technical staff, global engineering technologist at Dell. Welcome guys.

Bob Ward

>> Glad to be here. Back in the queue, man. Excited.

Rob Sonders

>> Yes, we are.

Bob Ward

>> Yeah. So a lot happening and I know in your space in this relationship, a lot has happened in a year. We've had some big announcements this week. Maybe let's start there. Give us the lowdown.Rob, you've got some great stuff to announce.

Rob Sonders

>> Well, so we sat here a year ago and there were some things I couldn't mention, which Dell automation platform and the big announcement there. But now we're sitting here on Tuesday and Arthur announced about Dell Automation Studio, which is now really allowing customers to build even more automation on top of all our platforms using things like Azure Local also coming with Dell Private Cloud. Underneath the Dell Automation Platform umbrella and then we can take Studio and automate even further, maybe bringing in all that AI goodness. Tell me about SQL AI or SQL Server 2025.

Bob Ward

>> Well, I mean, I was here a year ago. We were just announcing the public preview of SQL Server 2025. That's our flagship database at Microsoft. A lot of people know it. In fact, you may not know this, but Dell is the largest install base of SQL server in the world.

Gemma Allen

>> Wow.

Bob Ward

>> So people trust SQL Server, they trust the Dell platform. And we announced at our public preview to have AI integration. Well, we've gone general available. We pick years for names of products like SQL Server, so we don't miss the name. So we lasted last November. And the key architecture for AI capabilities is integrating with things like embedding models for vector searching or chat completion models for AI agents. But we do it all inside the engine, the SQL server engine. And the great part that we partnership with Dell about is that we can connect to anything on premises.

Rob Sonders

>> Yes.

Bob Ward

>> I can connect to Azure Local, Dell, Nvidia Factory. We just got a framework built in so it doesn't matter where you're running these AI models, we can easily integrate with it. And then you've showed me with the automation suite, you could automate that together, right?

Rob Sonders

>> Yeah, because you mentioned that AI now is here, right? We talked about it for a long time, but now let's do it. Well, now we're doing it with our automation suites that we have. If we want SQL server in the mix, we want Azure local in the mix. Maybe we even want AKS, Kubernetes.

Bob Ward

>> Right, exactly.

Rob Sonders

>> Because a lot of AI runs with Kubernetes. So we can automate all this now and make it easy. It really is the easy button to deploy all of these different ecosystems wherever you want them. And we can hook them Azure local, yes. Maybe up to Azure. We can build a compliment wherever the customer we meet, we bring the AI to wherever the data is, right? This is the story.

Bob Ward

>> That's right.

Gemma Allen

>> So AI is here. It's happening, right? We hear that. Bring this to life for us though. Give us some true customer work cases, examples of how exactly this is happening in the field. Where are you seeing the opportunities play out?

Bob Ward

>> Healthcare is an industry that I see quite a bit that still needs to and a lot of cases want to run their entire workload, their data on premises or in a private cloud environment. And yesterday at this event, I showed a couple of good examples of that. One with like a clinic on the Edge where they're going to start running maybe using Dell technologies with SQL, with AI models, all run on an Edge device, think about running at a clinic. The second example is like a hospital system, totally Dell private cloud, running maybe Nvidia AI factor with AI models, running maybe SQL server as devices and hospital rooms and so forth. So we can easily put this together. And the key thing for us is doing it securely. I'm sure you've seen this before, that running AI securely with data is paramount. So if I can control it on prem, first of all, that's great because I have full control over it. But for us, we use the SQL Server engine as that security blanket to access data with AI in a very, very secure encrypted way. So healthcare is certainly a big one, finance, manufacturing. We've already had several customers put this combination together with those industry examples.

Rob Sonders

>> You mentioned healthcare, manufacturing. Well, that falls right into everything we're bringing with Dell automation platform, SQL server, Dell distributed private cloud, another announcement today. That was the product formerly known as Native Edge. So we can put that anywhere in a hospital, central location, SQL server, running on Azure local, all on premises. But that's nice that Rob said all that, but let's make this actually happen with a blueprint, with a deployment, and make it all easy to manage day one and day two.

John Furrier

>> So connect to DoS. So Michael Dell, then Jeff Clark again here on stage today on his keynote, he talks about the operating model, this AI operating model. The automation platform is long game play. See that in the announcement. So if you're going to be an operating model, how does this stuff translate to the autonomous side of the things, the self-driving enterprise, whatever word you want, because the infrastructure's now got agents. You mentioned some of the examples. How does that kind of narrative operating model fit into what you guys are doing?

Rob Sonders

>> Well, I mean, we can put really, the agents, as close to the data as we want or where the data's being generated and that's maybe sourced back again to SQL Server if we need to get back to a central repository and that's the beauty of what we bring with all of that automation.

John Furrier

>> Connecting SQL to the system.

Rob Sonders

>> Yes.

Bob Ward

>> Absolutely.

Rob Sonders

>> And then we have partners that come over the top and maybe want to build custom solutions on top of these AI systems that we're deploying at an Edge, near Edge, far Edge, and then back to a central private location. Give some use cases.

Bob Ward

>> Yeah. I mean, again, the hospital's a perfect one for me because of the fact that they want to run it all in the private cloud, but they want to do agentic operations of these hospitals. For example, in a patient room, they want to be able to go into a digital chart using SQL Server as a data source and say, "I need clinical recommendations." And that's going to power off some sort of job that may go submit an order within the hospital itself, but using AI technology behind the scene to go make those decisions with the data. So anything you think about that is task oriented, anything that can be asynchronously submitted, but it's got to be secure with your data, that's what we're seeing and that's what we're using AI agents for.

Gemma Allen

>> So when we think about the why, right? When we think about the decisions that are happening now across IT decision makers and industry, there's a lot happening in a very short space of time and sovereignty and control is definitely a big, big part of this conversation, right? So we think about control planes, data planes, the kind of total cost of ownership element to that, the convenience element to that. What are you guys actually hearing and seeing from customers in the field? Like what is really driving the decision making, for example, to use Azure versus using just a self-managed Dell private cloud?

Rob Sonders

>> Well, let's use Dell Private Cloud to deploy Azure local on premises. That's the sovereign part of it in country, in region, right? Then manage it through the common control plane, which would be the hooks to Azure through Azure Arc.

Bob Ward

>> Yeah, Azure Arc. I think Azure Arc is one of the key pieces for us at Microsoft and this control plane idea because in Azure, you have the control plane built within the data center itself publicly in a public cloud. But if you want to Azure local, how am I going to actually still be able to connect myself in a way that makes sense to manage multiple Azure locals, for example?

Rob Sonders

>> But we also have to life cycle that hardware too. So that's where Dell Private Cloud maintains that compatibility and gives customers the confidence they can keep moving forward because you don't want too much drift. If you're in the public cloud, that's done for you. If you're in a private space, we do it via our day two operations.

Bob Ward

>> Yeah. Again, to me, Arc is a beautiful part of the control plane because it allows me to manage multiple different Azure resources, whether that's running in a public cloud or Azure local. I get connected to be able to look at billing information. I get connected to get updates so I don't have to do patching anymore. And then for us, the data plan is SQL server. I'm in a well-known database engine that just is powering not just your traditional relational database models, but again, I'm using it to do agentic experiences using the power of the TSQL language, but everybody kind of knows, right? So if I use that as a data plane and I'm connected to private secure AI models running on Dell platforms, I've got Azure Arc to go connect Azure local. I have both those things. I have both control plan and data plans.

John Furrier

>> That's a perfect scenario for sovereignty because...

Bob Ward

>> Absolutely.

John Furrier

>> You can now cross domains, boundaries, multi-cloud, whatever environment, so that's a key element. How does that play out from a deployment standpoint? Take us through like, okay, how does that work? You mentioned Kubernetes, obviously cloud native.

Bob Ward

>> Yeah.

Rob Sonders

>> So first we got for the on-premises solution, we're going to use Azure local, Dell private cloud, deploy that on-premises. The beauty of what that does is we start at the tin. We start in the box. We're maintaining hardware, firmware updates, rolling that through. Once we get to a point, then we connect to the cloud control plane. Azure bits come down, SQL server can be loaded through Azure. We could do it even with a custom blueprint if it's a specific need, however we want to do that. However, customers want to do that, but that's the full life cycle for that type of deployment.

Bob Ward

>> Yeah. I mean, using the automation combination with what local provides, since it is Azure in a box in a way, right, you have that complete autonomous of deploying virtual machines, Kubernetes services, AI models, SQL servers, Linux, Windows, all of that just comes together just like I was doing myself in the public cloud.

Rob Sonders

>> SQL server running in a container and Kubernetes also, right?

Bob Ward

>> Correct.

Rob Sonders

>> We talked about this many, many years ago. We did, and that was in some ways a bridge too far, but now it's commonplace because if you have AI, you have Kubernetes somewhere so now everybody's comfortable with it.

John Furrier

>> So the storage and computer separate. And so you can move AI to the data.

Rob Sonders

>> Yes.

John Furrier

>> And as they said on stage, that becomes one of the key value propositions.

Rob Sonders

>> Yes, yes.

Bob Ward

>> I kind of look at... You bring your data to AI because I think people are looking in terms of if they're forced to go to public cloud, they're thinking about moving their data up to AI technologies running there, which is great. We can do that with Azure, right? But if I want to run AI locally, why not just have my data with AI just coupled near each other, right? And the architectures we're building with both SQL and Azure local, which is really nice. We have customers in both worlds. They're in public cloud, they're on prem. It's the same. It's the same architecture, it's the same code base. All of it runs similar. So if somebody's going to deploy both, they don't have to learn two different skills to do it.

Rob Sonders

>> And Bob, just to, obviously we've both been using SQL Server. The ecosystem and the... So many people use SQL Server and they've been using it forever and there's such a community around this. And now having AI right there, it's a beautiful place to have all...

Bob Ward

>> It's interesting because I've been talking about AI now for a couple of years. A lot of people want to know, "How do I turn it off? I don't want to use it. I'm very scared of it." But now that I've started to teach them how to use SQL Server as a trusted data source with it together, they're starting to kind of get it a little bit familiar with it and understand the limitations of when to use it and when not to use it, especially if you control the models that you use. I think about Dell and AI factory is a really good example. Using like Nvidia models and Nvidia hardware, those are trusted brands and names that are people used to. It's not like I'm just going to hugging face and pulling down a model today, right? So once I start bringing in the more enterprise feel for this, people get a little bit more open to discussing it.

John Furrier

>> You get a large install base with Dell. Now the theme here is connect the data. That's step one. How does that work? That's in the platform, that's local?

Bob Ward

>> Yeah, it can be. Like SQL Server itself is a good example that connect to your data. If that's running on Azure local, could be in a virtual machine, it could be in Kubernetes services, but I'm able to automatically connect it with like Nvidia factory as an example because all our AI technology is just rest based. We're just talking rest to AI hosting systems to exchange data or do imprinting or embedding or vector search. So it's very well-connected, it's easy to do and it's all just in a private cloud and then you can automate it with your technology.

Rob Sonders

>> We also have to mention AI factory. You heard this also, right? It could be another... These environments could be another source for AI factor...

Bob Ward

>> Very easily. Absolutely. All connected on your private network.

Gemma Allen

>> So when we think about bringing AI to your data or not, right, this kind of mixed message right now around what the world of on prem looks like and what it would look like five years from now. We know that it's going to be largely decentralized, right? We know that one thing is potentially it's not a good idea anymore to keep your data in one place. When we think about sovereign AI and AI on the Edge and all of these different longer tail opportunities to enterprise that we weren't so excited about maybe five years ago or terms that it didn't even exist, they're now very opportunistic. What are you guys seeing there in terms of like, especially from the perspective of Azure and Dell and this partnership, how are those opportunities playing out in the field? Are these new customers... What is that TAM for you like?

Bob Ward

>> I mean, that's why hybrid is still to me a key technology that we work on together because I don't have a single customer that's running just on Dell on prem anymore of a SQL server. They're running both SQL and Azure. They're running now SQL and our Microsoft fabric. They're running it totally on prem and I need this consistent way of doing it, but I need a connected way of doing it. And to me, that's why Arc is a technology that allows me to still make that bridge. So as long as I've got that type of technology like Arc or Local that's consistent as a cloud, now I can expand and do what I want. I can shift and decide. Maybe I've got a couple customers who may decide to go, "I'm in a public cloud. I'm going to come back on prem."

Rob Sonders

>> It's a choice. We give choice.

Bob Ward

>> But the experience is consistent now, right?

Rob Sonders

>> Yes.

Bob Ward

>> It's flexible and consistent. Because of the fact I've got that right architecture, Azure Local being just running Azure, SQL running in any of these environments, I can pick and choose what I want and the experience and development experience is the same.

Rob Sonders

>> Yes.

John Furrier

>> Yes.

Gemma Allen

>> We hear a lot about data gravity, right? That's another huge topic right now. Again, seems like somewhat of a buzz term. What again, what is your experience, your field understanding of how, again, that is a central part of this whole conversation, right? We've talked a lot here at Dell Tech rather this week about memory, about talks, about all of these key parts of this formula to make these systems run as quickly and as instantaneous as they need to. But we know that there are challenges, right? Data has always been a struggle. I have a love hate relationship with SQL from my time at Microsoft I can tell you, but maybe kind of give us kind of your thoughts on that.

Bob Ward

>> Yeah. For me, you said love-hate with SQL, but one of the great things about it's been pretty consistent experience to go use the database itself. The security model is very consistent in the same. The performance model is the same and the availability model is very, very flexible. The fact that I can combine AI now inside the engine itself, I can do AI technologies with something I'm familiar with, that opens the ability for me not to have to go move all this data all the time anymore because what before we had this AI technology built in, people had to go take the SQL data, go move it somewhere else, do AI processing, bring it back in a SQL.

Gemma Allen

>> Find the data.

Bob Ward

>> We don't have to do that anymore. We don't do it. We can do it all within the same engine, the same TSQL interface they know and love and not have to actually move the data around. So just pick your model, run it in the right place and keep your data in a database that you know and trust.

John Furrier

>> All right guys, great job. I mean, this AI factory, distributed computing is here and it's just getting started. It feels like we're in this cloud on prem connected environment, fully hybrid. That's done deal. You mentioned Kubernetes. Then you got the Edge coming, seeing what's happening with the hyper converged Edge, more AI intelligence is getting injected in.

Rob Sonders

>> Everything is moving faster and faster. I'm having so much. I was telling Bob when we were standing over there, I'm having so much fun, the most fun I've had in probably 25 years right now with all the tools that we just really have at our fingertips. Using AI for all the things, my own programming assistant who works 24-7. So it works a little bit longer than me, not that much.

Bob Ward

>> The biggest difference between last year and this year's interview is the use of AI technology to go do all these things, to deploy the technology, to manage it and so forth. That's just going to keep going. We're going to see that...

John Furrier

>> What's next?

Bob Ward

>> Yeah.

John Furrier

>> What's this year? What's going to happen next year? Next year, it's a preview. What are you guys working on? What are the key things you're thinking about? What are you optimizing for?

Bob Ward

>> For example, here's one thing that we're thinking about is that how do we use agent technology to help people just manage databases? Simple stuff like that. Because today you can go into Claude Desktop or something and say, "Help me go tune a query." Could we put that inside our engine somehow to help people keep these more efficient? I've got customers running Dell systems that are managing hundreds of SQL servers and they're struggling to do it. It's hard. It's still good. But can I use agents to help me go do those kind of tasks?

John Furrier

>> I mean, I think SQL, I mean, you love it, but at the same time you say, "Only got to hire a SQL person to do it." I mean, it's a language It's a query language and it's Lingua Franqua for data, you can just abstract away and put intelligence on and make a AI builder or agent, digital worker, manage it so you get the best of both worlds.

Rob Sonders

>> Well, we have...

Bob Ward

>> Same thing for you guys.

Rob Sonders

>> Yes. We have AI, our agentic layer with AI ops that we're putting over the top of all of those platform automation platform solutions, everything is going to be part of that to look and literally just chat with an agent about storage or compute or what we need to do to make this more performant, add more, take away. It's all dynamically. So where's the human anymore, right? I guess occasionally just checking in.

John Furrier

>> Yeah, and opportunities to do some exploration.

Rob Sonders

>> Yes.

John Furrier

>> What are better queries? What are better prompts?

Rob Sonders

>> Yes.

John Furrier

>> So a lot of cool stuff hiding in the data.

Rob Sonders

>> Yes, absolutely.

Gemma Allen

>> Well, Robert, Bob, one thing is for sure, there is so much opportunity right now to really help enterprises drive value. So can't wait to see what you guys are talking about when we're back here hopefully in 2027.

Bob Ward

>> Yeah.

Rob Sonders

>> Thank you. Appreciate it. That was fun.

Gemma Allen

>> Thanks so much for joining us.

Bob Ward

>> Thank you.

Gemma Allen

>> I'm Gemma Allen joined by John Furrier. We are on the ground here in Las Vegas covering Dell Technology World 2026 where all the live action is happening, some great conversations coming up. Stay tuned.