Cisco Connects AI, Networking and Collaboration | Cisco Live 2026
Alan Shimel talks with Anurag Dhingra of Cisco at Cisco Live about the convergence of enterprise connectivity, collaboration and AI. Dhingra explains how Cisco is evolving Webex, workplace networking, security and Cloud Control so humans and AI agents can work together to manage infrastructure more safely and efficiently. The conversation also explores agentic operations, agentic security, scaling networks for AI, marketplace extensibility and why enterprises need an autonomy dial that keeps humans in control while moving at AI speed.
Transcript
Hey everyone. Welcome to our Techstrong TV coverage here at Cisco Live in Mandalay Bay in Vegas. I had to remember what hotel we were in for a second.
I want to introduce you to my next guest. His name is Anurag Dhingra. I got it right?
Yes, you did. Okay. And Anurag, first of all, welcome to Techstrong TV.
It's great having you up here. Thank you for having me, Alan. My pleasure.
I'm looking forward to this conversation. I am as well. Anurag, we should start with your title.
They're going to see it on the bottom third of our screen. Right. But for those who aren't reading or...
Yeah. Tell us what you do at Cisco. So I'm a senior vice president and general manager for Enterprise Connectivity and Collaboration.
Long title. It's a long title. Enterprise Connectivity is really around networking in the workplace, your Wi-Fi, your switching and routing.
And collaboration, of course, meetings and messaging and new video conferencing. That's really the scope of the role here. But that covers up a whole lot of territory- Yeah ...
you just mentioned there. We're going to dive into it in a second. But before we do, I want you to give people a sense of your own journey, your own path.
Yeah. SVP, this is a good- Right ... sizable chunk of Cisco's business.
Yes. Yeah. How did you arrive here?
Yeah. So I'm an engineer by training. Okay.
And when I started my career, I started as a software engineer. I was writing code, and I actually started at Cisco, this was two decades ago. So I've been here for a long time.
I started as software engineer. Over time, I took on more leadership roles, but I worked in the networking space at Cisco. I also took a role in security at Cisco.
Oh, really? I ended up building products in the collaboration space. So that has been my career journey at Cisco, and now it's kind of really a privilege to manage a team that is focused on the intersection of connectivity and collaboration.
And so I'm really excited by that. I will tell you, I've been in tech 30 plus years myself. Right.
It's not often we see someone with your title who really rose up from the ranks of actually writing code. I know a lot of people who made the switch from coding to business and have done well. Right.
Actually, someone probably you know, Aruna Ravichander. Absolutely, yes. Aruna- She's actually my partner in crime.
She's the marketing leader for my business. Really? I know Aruna many years.
Yes. And her and Jeff, great people. Yeah.
But she was a coder with hundreds of lines of code. Yes. In the product and made that switch.
Right. As you sit here today, you described briefly about a little bit. When we say collaboration in Cisco, a lot of it is the WebEx brand.
It is. And they invented video conferencing- Yeah. Right ...
for most of us. Yes. But today's WebEx isn't the WebEx it's 10 years ago.
It's not. Talk to us a little bit about that. Yeah.
So I think, look, when you look at where we were and where we've come, we basically rebuilt the product from the ground up for two reasons. One, of course, technology changes over time. Mm-hmm.
And so you have to keep up with that. But second, the way we collaborate with each other has evolved quite a lot. It used to be phone calls in the past, then it became meetings with people, then it became a very messaging first world, and we are in some ways living in a very messaging first world.
And so we had to evolve the product to go way beyond what original WebEx did. And these days it's all about not just human collaboration, but humans and AI working together. So the product had to evolve.
So it's a very, very different product as a result of that. I'm going to come back to that in a little bit. Yeah.
But first, I want to give equal access to the networking piece, right? Of course. Today's offices, when they have an office- Right ...
are a hybrid. No one runs Cat 5, Cat 6. Right.
It seems everything is wireless, and sometimes I ask some of the, I call them kids, some of the younger people in my company- Right ... " Right. What percentage of that traditional...
Right. Cisco was switches. Yeah.
Right. Cisco was ethernet. Yeah.
How has that business changed in Cisco? Yeah. So that business has also evolved, like you said.
When we think about people connecting to the network, it is a Wi-Fi first world. There is no question, right? In a workplace, it is usually Wi-Fi.
Outside, it's maybe 5G connectivity. It's a mobile first, Wi-Fi first, wireless first world. However, when you think about where those wireless access points are connecting, they're still connecting through wires to a switch somewhere.
Somewhere there's got to be a wire. And then if you look at a modern building, there is actually many more IoT devices there. Yes.
There's sensors, there's cameras. There are many more of those than actually laptops and mobile phones connecting to the network. So the connectivity for many of those things is still wired, and so it's just moved from what is connecting to the network, and not really a question of only people, it's people and things.
And so the relevancy of switches and routers and wireless is more than ever now. And with AI agents, that has to be looked at again because as agents start to do work like us, they're running on our laptops, they're using tools that we use. They're connecting to the same network.
We have to make sure that infrastructure actually scales. The agents never sleep. No.
They don't take a break. They're constantly running, so the network never becomes idle. Has to be up and easily.
Exactly. The network is never idle as a result of that. Well, but that creates its own set of problems, doesn't it?
Absolutely. And that's one of the things I think we're learning about this whole AI thing, is AI scale Yeah. It is 24/7.
Right. It is wider, broader, faster. Yeah.
And if we don't have systems that are set up to handle that scale, you're going to have failure. Yes. Yeah.
So here at Cisco Live, we're talking about three themes here for how you actually scale AI. The first is you have to scale the network itself. Yes.
And we obviously build devices, routers, switches, router, wireless, all of that has to scale up. The second is concern around security, securing the infrastructure, and the new threats that are emerging because of AI. How do you secure the agents themselves?
So that is a core concern. And then the third thing is very real for operators, the IT teams that manage this all. Complexity is through the roof, and at the same time, there's just not enough networking skilled people in the workforce right now.
Yeah. And so we are looking at this holistically. How do you scale the capacity of the infrastructure?
How do you secure it? How do you make it easy to manage? And AI can be an answer in all of those areas.
So that's the good AI. Yes. Now of course there's the bad AI, right?
We hear about mythos and- Yes ... vulnerability apocalypse and all of these things. You're sitting here between your collaboration and your networking.
We've got millions of lines of code out here. Right. Does it keep you up at night?
Absolutely. But I'm also optimistic in this sort of new world, right? And the reason I'm optimistic is I think some of the companies, whether that's Anthropic or OpenAI, they have chosen to do the responsible thing where they're giving the defenders a sort of a running start.
Yes. Before these models that are really good at finding cybersecurity risks, before they become released to the world, they are accessible to us and many of the other companies that run the critical infrastructure of the world. So that gives us a head start so we can use the models to find vulnerabilities, patch them before the attacks start to show up.
But the one thing that we have to do is we have to make sure that we are ready for this constantly evolving world. So it's not a one and done. No.
We have to stay on our toes. But it's an exciting time that we have access to such powerful models that can help us protect the world as well. I tell people all the time, we're living in a historic time.
Right. I got involved in technology a little before the dot com. Yeah.
But I sold the company in the dot com, helped take another company public. Yeah. I lived through that whole thing, up and down.
Right. And this is so much more than even that was that. Yeah.
It really is crazy. I want to turn now specifically to Cisco Live this week. Right.
We saw the keynotes today. Lots of great innovation. No doubt Cisco is creating...
" Right. Right? Specifically as it relates to the product lines that you're responsible for.
Yes. What's coming out of Cisco Live here that our people at home should be aware of? So I think one big announcement we're making here, is Cisco Cloud Control.
Yes. And that is very applicable to every product in Cisco, of course, my business line as well. And Cloud Control is the front door to our platform.
It is the place where humans and AI agents can work together to manage your infrastructure. So that's a very exciting announcement. It's very applicable to campus and branch networking, very applicable to collaboration products as the rest of it.
So I'm very excited about that, as there's a lot of interesting technology there. But like I said earlier, we are focused on really three top of mind concerns. Scaling networks, so we're announcing brand new switches and routers and access points to scale networks for AI.
We are making announcements around agentic security and how networks actually become a critical infrastructure piece for enforcing security. When you think about it, everything connects to the network. Yes.
Everyone connects to the network. Agents connect to the network. So it's a really interesting control point for applying policy, and so I'm very excited about what we're doing there.
And then using agentic operations, using AI agents to simplify operations massively. There's a lot of interesting capabilities that directly apply to managing collaboration infrastructure as well as campus and branch networking infrastructure. So I'm very excited about all three of those areas.
I can tell it's coming through, and that's great. Let's talk about the... It's Cloud Control- Yes ...
product. As someone sitting in the audience, I don't have the insight you do. Yes.
It really, this is, to me, the first step here, because this is a platform play. Yes. And now you need the partners and the industry to layer in, right?
Yeah. Eventually, I think you'll see a marketplace, if you will, of solutions- Yes ... that play on top of this.
It's also open sourced or a lot... I'm not going to call things that are open sourced. Some of the technologies, yes.
But a lot of it isn't, and it gives everyone a chance to... So is that the fabric of the OS for the AI infrastructure there? Yeah.
I think you can call it the new operating model for- Okay ... how you manage infrastructure, right? As a product, it has capabilities that let you manage everything that Cisco delivers to our customers, whether that is networking or collaboration, security and observability.
All of our products can be managed in one place. Like you said, it is extensible, so third parties can extend the platform. Right.
And we announced a marketplace with 50 partners. I think 50 plus partners to start. 50 plus partners.
But the second piece that is interesting is really how agentic capabilities become available to manage your infrastructure. And Cloud Control gives you that explainability, the auditability, how you supervise and manage agents-So it is really a place where a blended team of IT operators, humans, and agents come together. So it's a very, very exciting start of a new paradigm for operating models.
I agree it is. It came up in, I think, the press conference after the keynotes today. Look, a lot of people are now starting to say agentic is entering a reckoning on finance.
Yeah. Similar to cloud and FinOps. Right.
Yes. Are you worried about the token cost of the agents that are here? Right.
Yeah. I think that's a top-of-mind concern for everyone. In the industry, there is this term that people use, it's called agent maxing.
Yes. And the idea really is you're relying more and more on agents, and of course, there's the real cost of that. So I think as an industry, things will become better.
I think industries have, technologies have a way of optimizing themselves. At the end of the day, the technology has to deliver real ROI. Yes.
And that only happens if the economics work out. And we are playing our part in it. The industry as a whole, I think, will figure out how to solve this problem.
For now, I think just getting visibility into that token spend is really, really interesting. And then we have capabilities in the cloud control platform. We call it tokenomics.
So you can at least start by looking at what you're doing, and then you can start to then start thinking about optimizing that. I want you to say that word again, because it's the first time I've heard it. Talk to that camera.
Tokenomics? Tokenomics. Token- Token economics.
So- There you go, tokenomics Right. You heard it here first. Yes.
But no, look, I will tell you, guilty. As the CEO of my company, I literally forced everyone- Right ... to adopt AI and agents.
I said, "Don't worry about the cost. " Right. But we had an incident this week that we had 5,000 videos that were already uploaded on one platform.
An agent was trying to upload them to the same platform. Right. And our meter was running, because we get charged by how many videos got uploaded.
And I said, "But those videos are already up there. Why are we-" Yes ... " Right.
" But this is the learning. Yeah. I think you're touching upon how do you make sure that the agents do the rational, logical thing.
Right. That's it. Which is touching upon the safety and trust aspect of that.
And honestly, these agents are powered by models that are non-deterministic, so this is not an easy problem to solve. However, some of the tools that we're announcing here at Cisco Live, both on the tokenomics side, but also in terms of just agentic behavior, agentic security, guardrails, I think they are part of the solution here. But you're absolutely right.
I think these are top-of-mind concerns for our customers. The way I look at this is, the autonomy is a scale, it's a dial. We want to make sure our customers have the ability of not just supervising these agents and feel comfortable in the reasoning and the evidence that the agents are using, but then you have control of how far you want to go in that journey.
How much control you want to retain and how much control you want to delegate is a critical part of the puzzle, and that is the responsible way to do this, versus here's a black box, trust me, it's all going to work out. That is a recipe for disaster. Yeah, trust me, yeah.
So we're squarely in the first quadrant. I agree with you. Last question, and we'll let you go.
Put Cisco Live behind us. Yes. We're in a state of technology right now where it seems things are changing with every release that a frontier model comes out with.
Right. What does it look like six months from now, a year from now? Right.
I think predicting the future is extremely challenging. That's why I kept it to six months to a year. So even six months- Yeah ...
might be an eternity in this day and age. No, that's how crazy it is. So here's what I would say.
I think the potential is real with agentic AI, but I think agents maximize the opportunity, and at the same time, also maximize the risk. So I think what will still be relevant in six months is, can you scale your infrastructure? Yeah.
Can you deliver this at security and safety? Yes. And can you actually manage all the complexity, simplifying the operation?
I think those three themes are durable. I think they're going to last that long. And I think at Cisco, our role is very clear.
We are in the business of making sure our customers can run at the speed of AI and do that while maintaining control and safety and security. So I think that's really what we are focused on. Those are the products that we're building.
This is something I hear continuously. You need to go to AI scale. You've got to do it with governance, guardrails, security, whatever you want to call it.
The third thing I hear, though, is you got to keep humans in the loop here. Absolutely. Right?
Yes. We're not going to go to just a totally agentic enterprise. Yeah.
I think this is why I talk about that autonomy dial. I think keeping humans in the loop is very important, and giving our customers control when they feel comfortable and how comfortable they are in terms of delegating control is really critical. And you will see that from all of the products that we're announcing here at Cisco Live.
Cloud control becomes the- Yes ... center point, the front door to that platform. But safety and trust is a core principle that we're building towards.
Love it. Thank you so much. I appreciate it.
Thank you, Alan. Hey, I hope you've enjoyed this little peek into what's happening here at Cisco Live. We've got more interviews.
They'll all be on Techstrong TV as well. You can check them out there. But for now, we're out.