Cisco Reimagines Customer Experience for AI Era | Cisco Live 2026

Hi, everyone. Welcome back here to our Techstrong TV coverage of Cisco Live. We've had just an absolute power lineup of speakers today, and I think this is the culmination of it.

Let me introduce you to Liz Centoni. Liz, first of all, welcome to Techstrong TV. I know it's your first time on here.

Thank you for coming on. Thank you for having me. I'm excited to be here.

Well, it's my pleasure. Let me let you introduce yourself, Liz. If you wouldn't mind, tell our audience a little bit about you, your current role, maybe a little bit of your history leading up to it.

Okay. So I run our customer experience organization. I've been at Cisco for 26 years.

Twenty-four out of those have been in product. Really? I started in Cisco in 2000, so I've been in engineering a good chunk of my time.

Then ran a few businesses, IoT, compute, our applications business, and then also led the acquisition of Splunk. So my move into our customer experience organization has been recent, just two years. Yeah.

And here's, I think the way I would summarize it is, throughout my career, it's always been about, are we giving customers what they came to us for? And I think that's the same thing that happens that is about our customer experience organization. Because a big focus of my team is taking what the product teams have built and helping our customers deploy it in their organizations and getting the most value out of it.

" And that pleasant surprise is what we strive for every day. So the motto for my team is always put the customer at the center of everything that we do. So it's not just about building technology, building capabilities for us, it's about being able to deliver that mission, putting the customer at the center of everything we do.

I love it. What a refreshing outlook on what customer experience should be. Unfortunately, or fortunately, I've been involved in several organizations where we had customer experience teams, help desk and so forth.

And I got to tell you, it wasn't always as eloquent, simple, and to the point like that. I used to think somehow we were just in the bad news generation business. It was how do I escalate from level one to level two, and level two to level three?

But I think what you've enunciated here is the real goal of what customer experience should be. Delivering delight- Mm-hmm ... is what we used to call it.

We have a lot to talk about, but one question I wanted to ask you was, in this age of AI, so many companies seem to be abdicating their customer experience opportunity to please a customer by putting in a chatbot, by bringing the AI and taking the human almost out of the equation, making it almost impossible to get a real person. How do you, not specifically even to Cisco, but as someone who runs one of the biggest customer experience programs in the world, how do you view that type of automation? We started there as well.

Oh, yeah? I would say. We had chatbots.

We did a lot of automation. Mm-hmm. The goal and the intent was absolutely there in terms of serving our customers.

How we went about it, initially to where we are right now has changed, and let me give you an example of that. " It's like, "Don't you have all of my information? " Right.

They have to give it to you again. And we're thinking, "Great. This is perfect.

" Right? AI does summarizations really well, so the way to look at it is make sure that every engineer that it gets handed off to has that summarization, so they're never asking the customer that question again. Well, what it did was it just made our customers even more upset faster because what it surfaced to them is that, boy, instead of getting the right expert, I was handed off to four different people.

So it's an issue. So we rethought this, and we said it's not about bolting AI onto a process that was already broken because all it did was it just made the process run faster. Worse.

AI did the job that we asked you to. We went back and redesigned the work itself, which means you have to get close to the work. Yeah.

You can redesign it sitting at the top. And our intelligent routing now routes the cases close to 90% of the time to the right engineer the first time because the metric that we used to measure ourselves initially was around summarizations before they had making the handoffs much worse. This was really around first contact resolution.

And so again, we had to pivot from what we thought was the outcome that we're going for to the outcome for the customer, and that was the unlock moment for us. Yeah. And it's one example of it.

And here's what I would say is, sure, we're using AI in terms of automating some of those repeatable tasks, but there are things around where we need those human kind of the relationships, the judgment that's needed. Many times we don't have complete data, we don't have complete context. We don't use AI for that.

And for us, this is where customers are deploying it in a production environment. So we're takingA model that's probabilistic, and the outcome- Deterministic ... " Try this instead.

It's a one environment. Oh, you were right. Let's-- Yeah.

Exactly. I love it. That focus, and so the human's always in the loop for that, and we have very high thresholds.

And when those thresholds are not met, it deflects to a human. That's excellent. A lesson there for everyone in customer experience.

Liz, if you don't mind, I have some notes I want to jump in on. Look, this has been a jam-packed Cisco Live in terms of releases and new information because let's face it, we're in the dawn of a new era. And Cisco's not staying left behind.

They've come into this with both feet. If you don't mind, I want to hit a few of our things, a few of the announcements and news that came out of this. Resilient Infrastructure Services.

Tell us a little bit about that. So we went general availability with Cisco IQ, which is the delivery vehicle for our AI-powered intelligence for our customers who buy our support services and our professional services. We went GA about five weeks ago.

As of 8:30 this morning, we had 2,036 customers on it. Cisco. And mostly customers are onboarding.

" And the answer is pretty simple. This is the first time our customers, for their Cisco environment, like hardware assets, software assets, cryptographic assets, for when they have an always-on dynamic view of everything in their environment. And if you look at what's happening post-Mythos is our customers need to know what's in my environment, what's supported, what's not supported, what went last day of support sometime back, what are the vulnerable, and what is the security posture?

Because then I can then put plans in place, which my team will help with resilient infrastructure services to help them, one, identify the risk. Two, help mitigate the risk, whether that's patching or it's refreshing the hardware. And there's a third part of it in terms of going, Mythos is not stopping.

That's just the first of many models. The next generation of AI-generated attacks will come about, and this is scary, where you can map your environment in minutes and find those vulnerable devices. So we want to help our customers with continuous assessment around the networking and security resiliency of their environment.

I would love to say that 18 months ago, we had this foresight that Mythos was coming out. Yeah. And hence we built Cisco IQ.

We didn't. We built it for a different kind of generation of attacks, like SolTyphoon. It was perfect in terms of timing around helping our customers with this three-phased agentic approach.

This is just the beginning. We will, along with our customers, continue to iterate and evolve this because no one has fully figured out how to address this. I learned a lesson a long time ago in business from a mentor of mine.

Sometimes it's better to be lucky than smart. It's good to be both. But you're right.

But I couldn't help, this is a theme that I saw around the experience announcements at Cisco Live this year is that they were very security focused. Everything from, as you said, Project Mythos, Glasswing, post-quantum security. As a security person for 25-plus years, it did my heart good to see security taking its rightful place up here.

It's important, and it's not just a silo. I was talking to Peter Banks earlier. We were comparing notes.

He's been in security a minute or two as well. I think finally, people are realizing security, it runs the whole gamut. It's embedded into everything, especially a customer experience program.

Absolutely. And security fused into the network. Mm-hmm.

You will hear us talking about it. You can't just treat networking and security in silos anymore. Anymore.

Yeah. That's the old way of looking at it. It's almost like saying, if you're looking at it the old way, I'd say 2008 called and wants it back.

It wants your security back. And I hated 2008 security. I mentioned quantum.

I want to talk a little bit more about that. For as long as I've been around, people say quantum, five to 10 years out. Mm-hmm.

Nuclear fusion, five to 10 years out. Well, it's not five to 10 years out anymore. It's literally on our doorstep.

'28, '29, sometime then, we're probably going to have Q-day. Of course, no one wants to be caught sleeping at the wheel. And again, your organization is jumping out in front in promoting quantum readiness assessments.

Mm-hmm. What exactly are you doing with that? Yeah.

So as you said, it's now. Yeah. It's not some time in the future.

And there's regulatory pressure as well in terms of addressing this. The reason that we did quantum readiness assessment for three reasons. One is to help our customers first identify and find cryptographic assets in their environment.

So identifying and just even understanding what do I have, not just hardware assets, software assets, but cryptographic assets as well. Second is the readiness assessment. When my team goes in and does what we call this quantum ready assessment for our customers, I would say in the majority of time, they're surprised at their readiness.

Just even one question when you get down to it and say your most sensitive dataAnd you ask the question, when you think about harvest now and decrypt later, what are you thinking when you need to keep this most sensitive data, like in banking and other industries that are regulated, for the next 10 years? How are you protecting this? How do you address the nation states that are grabbing all this information, hoping that at some point in time it'll be decrypted?

They're surprised at their lack of readiness. So one is we help them identify their assets, two, help them map their readiness. " Where you are.

Now let's tell you. And then a phase plan, because the environments that we lived in are mixed environments. You have your traditional networking, and then you have your quantum networks as well.

And that's going to continue for a period of time. So the technology is awesome in terms of, like when we talk about in healthcare, it's an area that's so near and dear to my heart. I think Chuck talked about this.

When you take this combination of quantum and AI, in terms of whether it's drug discovery, whether it's around- Cancer ... helping with cancer, which is something so near and dear, in cancer treatment, so near and dear to my heart. But then you look at the challenges that come with it.

So for me, on my team, the focus is just the assessments. Get customers to understand where they are, then help them with a phase plan, including with the portfolio that we're building. And this is where it's not a one and done, just like what we talked about with Mithos.

It's a continuous assessment. I think it really is a case, an ounce of, what's the old saying? An ounce of prevention is worth a pound of cure.

If you get out ahead of it now, there are a lot of post-quantum algorithms out there already. There's this whole program. You don't want to be the person sleeping when Q day comes.

Let me jump onto our next topic here, which was Cisco IQ on-prem. So in a past world, I sold a lot of security to the federal government. I remember the Department of Interior had schools for the Navajo children where you can only do line of sight connectivity.

Mm-hmm. And sometimes it worked, sometimes it didn't. So this on-prem, where you're not cloud connected, you're not network connected even, is an important piece of it.

Talk to us about this aspect and what we have here. So my engineering leader, Bhaskar, and I, we built Intersight. I would say kind of the first organic SaaS platform within Cisco, managing compute from the cloud.

We learnt our lesson at that time going, if you look across our customer base, there are customers who are fine with SaaS, there are customers, even outside of the federal government, who for whatever reason, sovereignty, want that data governance and control, and they want it on-prem. They want it on-prem tethered, they want it on-prem air gapped as well. So we decided to do something completely different than what we had done before.

Is that as soon as we announced the SaaS deployment mode, the on-prem deployment mode would come in three months later. It's exactly the same set of capabilities, the same intelligence, but it gives customers the flexibility of their deployment mode, their control, their rules. So we learned from our past experience- You learned from real world ...

that this was super important. You know what? Look, I've been on both sides of this one.

Really what we're talking about is trust. It is. And even though Cisco's a trusted brand, there are customers who just don't trust any...

It's the old story. How can I be totally secure? Just unplug from the internet.

That'll go a long way. And there are people who have those trust issues. So, it's great to hear that, and they're not a second-class citizen here now.

Right? Absolutely not. Fantastic.

Want to talk about peer benchmarking. It was one of the features that was the most asked for by our customers. This is even before Mithos came in.

What they said is, "My board asks me often, my senior management will ask me, 'So how do we compare? '" And they didn't want a generic report. They wanted the actual comparison.

So we provide, obviously, anonymized actual comparison within industry, within type of assets, within infrastructure size. And now with Mithos, they're asking the same question, going, "So, are we behind? Are we about the same as our peers?

" So this capability of doing peer benchmarking across all of this was one of the top features asked by our customers, and it's coming in next month. Well, I've been speaking to CISOs since before we called them CISOs. " Yeah.

It's the most important thing. Liz, I've got more to go over, but unfortunately, we're almost out of time. We're beyond out of time.

I would just want to say, first of all, I'm a fan. Second of all, congratulations. You're doing a great job with this experience program.

I think it's everything you'd want a company like Cisco to have in a customer experience program. Thank you. Thank you for coming on.

This has been wonderful. Thank you. I have to say, this is an effort of a lot of people.

It takes- It takes a village ... my team, it takes the sales teams, it takes our product teams, and all the other functions at Cisco. Nothing happens just by itself, and I'm glad to be able to do this at Cisco.

But good leadership always shows. Hey, we're going to take a break. You're watching TechStruck TV.

We're coming back with more Cisco Live.