Splunk Powers Cisco Cloud Control Data Fabric | Cisco Live 2026

Hi, everyone. Welcome back here to Techstrong TV's coverage at Cisco Live. I'm really happy to have my next guest here because it's a topic and company I've wanted to talk about for the, this is our third day here.

Let me introduce you to Mungish Pimpalkhare. Yes. I got that right?

Yes, you got that right, Alan. All right. I feel like I won already.

Welcome, Mungish. It's great to be here. Thank you for having me.

Great to have you here. Well, let's start off, your current position is with Splunk, and what is your title? Yes.

So I'm the Senior Vice President and General Manager for Splunk Platform. I love it. But let's talk before this role.

Give our audience a sense of kind of where you've been, how you got here. Sure thing. So I've been at Splunk for close to seven years now, and prior to that, going all the way back, where I very much started in the industry as an engineer, initially at semiconductor and infrastructure companies- Very cool ...

designing GPUs before they were called GPUs. Well, they were games. Yes.

They were for games. They were graphics controllers. Right.

Exactly. And then had a mid-career stint in venture capital for eight years. Good for you.

And then came back to the industry in product and general management roles. And, it's been very, very exciting now to be part at Cisco to kind of bring some of my learnings and really have the opportunity to amplify the impact at such a large scale. Absolutely.

Absolutely. And, it's funny you said GPUs. It's graphic accelerators.

Yeah. I remember up, back in the day, when you would update your graphics card, you'd had to go in and set the ICQs and the conflict things on your motherboard. Yes.

Oh, I don't miss those days, I got to be honest with you. But, Splunk, and we were talking off-camera, I followed Splunk, I think, since the beginning of Splunk. It's always been an amazing technology, and it was this just such an amazing technology that you could go off into a lot of different areas with it.

The Splunk that we see today really kind of plays in two areas, one observability and one in cyber. Mm-hmm. And there's an overlap between them, but they're two distinct missions as well.

Mm-hmm. Talk to me as, as VP of product- Mm-hmm ... and GM, how do you balance that, right?

Do I put too much in here, not enough there? It's a balancing act. Yes, it is a balancing act, but I'll say that the opportunity has never been greater to look at all the spectrum of use cases, spanning both security and observability, with a common foundation.

And, Alan, you've followed Splunk for a while. Going back, Splunk was always in the digital resilience space. Yes.

And if you go back to the history of Splunk, it actually started as a platform, a broad-purpose tool, and looking at machine data, and then it found a natural home in security use cases- Mm-hmm ... in CISO orgs, and IT operators, and DevOps has been using Splunk pretty extensively. And if you look at all the changes happening with AI now, what you can't observe, you can't hope to secure.

So in some ways, it all starts with common visibility across all the operational activity and looking at the data, all the way to end use cases. So that's how we try to balance out the different use cases, by looking at what's common and then pushing that down to a common platform. And then, of course, we have domain-specific capabilities for security or networking or IT operations.

I love it. It really was, because at its heart, what I remember about Splunk is it could, no matter what your log file, what format your log files were doing, if there was an information stream, Splunk was able to take it in. And then where I think Splunk really hit its stride was gathering that information was great, but providing analysis on top of that information- Yes ...

actionable intelligence, as we used to call it, right? Yeah. That's when the light bulb went off, right?

And then, if you want to take it from a security point of view, actionable, right? Yes. We didn't even call it observability then.

Let's be honest. Yes. Yeah.

But, you want to go down the ops point with it, it's just amazing tool. Now, Splunk's also a great example of a post-acquisition, new beginning, second life. Yes.

Look, you've been in VC, you know this a lot. Acquisitions fail sometimes, right? You got to break some eggs to make omelets, as they say.

Yeah. But Splunk, I think, represents a great success in that Cisco has allowed it to maintain its identity- Mm-hmm ... while integrating its functionality into this broader offering.

Yes. Again, a balancing act, right? There's another kind of balance act.

How do you, as the GM, maintain that balance? Yeah. So that's been the most exciting part, especially in the last couple of years.

Mm-hmm. That Splunk had the great opportunity of not only building on its value in the security and observability space, but now with Cisco's massive footprint in the networking domain, we can apply a lot of the same capabilities you talked about, Alan. Yes.

That applies as much to the networking space as it does to infrastructure applications and so on. So now, here's an opportunity to expand our rally, going beyond our traditional suites, part of specific domains like security or IT ops, or DevOps, and extend that to almost all operational teams within an enterprise. And we are, of course, starting with express intent to look at networking operations and a lot of what Cisco's doing with the overarching redefinition of operations.

Splunk plays a big role in that. Absolutely. I guess we should talk a little-- We're at Cisco Live.

Yes. We should talk a little bit about Cisco Live. Yeah.

Some interesting announcements, both Splunk specific as well as, again, Splunk as part of this broader ecosystem. If you wouldn't mind, share with our audience some of the announcements or some of the news that you think is important- Yes ... from Cisco Live.

So it's been really an exciting Cisco Live. And to just start off with, the big announcement here was, of course, Cisco Cloud Control. And for those who might not have heard about it, you can think of it really as the management and orchestration layer for driving agentic operations in an enterprise.

And it does many things beyond that, of course. First and foremost, it is a single point for having controlling and managing all your Cisco products. But then more importantly, it provides a framework for governing agents and kind of implementing agentic workflows, whether you're in security operations or IT operations, networking operations, so on.

Now, Alan, this kind of amazing solution requires the underlying data, the cross-domain telemetry, and that's where Splunk comes in. So what we have been doing is really kind of expanding Splunk's traditional architecture, and we call that Cisco Data Fabric, and it was purposely named Cisco Data Fabric because it's the next evolution powered by Splunk, and it backs up all the Cisco Cloud control use cases with that kind of cross-domain telemetry insights and all the knowledge context that is so important to making agents come to life. Absolutely.

It's great stuff. I wrote an article this morning, actually, it was published around cloud control. When you sit down and put pen to paper-- Well, no one puts pen to paper anymore.

But when you sit down to start writing the article, you realize how broad a mission this is. Sure. Splunk is part of it, but it's funny, Splunk plays in multiple parts of it, both as one would think, security and observability, but it's a really broad set.

The key to it, though, is it's a tool for the agentic age. Right. How is this whole agentic AI thing affecting Splunk?

Yeah. We are looking at agentic AI from really two angles, right? First of all, we are making extensive use of AI models and agents in our own solutions.

So making it smarter for customers to onboard data. We have, for example, AI-enabled data onboarding, so that now customer administrators don't have to spend a lot of manual toil tapping all the data pipelines that you talked about before. Sure.

And all this is largely automated, and even the continuous optimization of storing data in the right tier at the right time. So one example of that is we announced Machine Data Lake, as well as the concept of federated search and analytics. What that does is it gets customers away from having to overspend money into Splunk, without knowing what the use case is.

So with the modern Splunk architecture, now customers can just use AI to onboard data and keep it in the right data tier and think of then optimizing the cost to light up the right use cases at the right time. So that's an example of how we are using AI. And then to your other part of your question, how are we becoming more relevant for agents and agentic operations?

We're really building the next generation platform to power agentic use cases. Now, what does that mean? That means agents have to have AI-ready data, the right training models.

So it goes all the way from this distributed data architecture to purpose-built AI models, all the way to those AI-first experiences that you mentioned from Cisco Cloud Control. So here's the problem, not a problem, but here's a challenge that I see today. Is Splunk going to give me Splunk specific agents to use, or I'm already using, take my pick, OpenCloude, Claud Cowork, Nanocloud, this cloud, that cloud.

As where you sit, how do you support all of these different agents? Because they all could be bringing you information as well as acting off, they need to access information to act off of it. For sure.

So we have always had a philosophy of being an open, interoperable solution. Yep. As much as we want to give our customers great value out of the box, we very much want to continue to be an open, interoperable platform.

And in the context of AI models and agents, the answer to your question is both. So Splunk does supply domain specific models. For example, we have the AI toolkit, and part of that has been specific domain models such as the time series model, right?

That allows customers to make forecasting predictions on signals they're seeing from machine data. But more importantly, we also announced something called Agent Builder, and this allows customers to bring in their own agents, whether it's Cloud, Codex, or other agents, and then use the framework that Splunk supplies, that Cisco Cloud Control supplies with these external agents to work hand in hand with all the agents that come out of the box from Cisco. Love it.

Yeah. I love it. Anything else here at Cisco Live you think our audience might want to know?

No, it's been super exciting. It's really great to now see Splunk very well integrated into all things Cisco, and more particularly, as we continue to strengthen digital resilience, we really have a great opportunity to now become part of the entire Cisco landscape with Cisco Cloud Control and redefine operations in an enterprise. I love it.

Yeah. I want to thank you for coming on and sharing. I know it's a little crazy, but carving time out for us is much appreciated.

Thank you, Alan. Thanks for having me. My pleasure.

Hey, we're going to take a break. We're here still at Cisco Live. We've got a lot more coming.

It's almost lunchtime, but we've got, I think we're working through lunch. We got a whole afternoon. No rest for the weary.

You're watching Techstrong TV.