Cisco Webex Dialog Moves AI Customer Service Beyond Chatbots

Hi, everyone. Welcome back here to Techstrong TV. Continuing our coverage of the Cisco Webex One event here in Austin, Texas.

And we're at the Fairmont Hotel, which by the way is a beautiful hotel, right next to this big hole in the ground where they're building the next generation convention center for Austin. I kind of like the old convention center, but progress, as they say, right? Let me introduce you to my next guest.

His name is Vinod Muthukrishnan. That was fabulous. Better than I would do it.

Vinod, welcome to Techstrong TV. It's a pleasure to have you on here. No, thanks for having me.

So let's start. I'm going to let you introduce yourself to the audience, and maybe if you could tell them your title a little bit, maybe about your journey to your present position. So Vinod Muthukrishnan, I'm the VP and general manager for Webex Customer Experience.

I actually came to Cisco because I had a startup called CloudCherry, which did customer experience management. And the holy grail of that, the journey inside analytics, predictive analytics platform was, why do customers do business with the brand? Why do they come back?

And what drives loyalty, and what have you. Cisco was an investor in the company. And then when we got a couple acquisition offers, Cisco leaned in and said, "Hey, we're not in the contacts in the business.

" And I dealt with Cisco, as an investor, as a partner. I was very culturally aligned, and I felt this was the right home, for my technology, for my team, and also for me. So here I am.

Really? How long ago was that? Two lifetimes ago.

It was in 2019 when the world looked fairly different. Yeah, it was a very different world. It was pre-COVID then.

Pre-COVID. I did have a short stint away from Cisco, after COVID, but- So you went boomerang on top of everything else. I'm absolutely a boomerang.

And I think when Jeetu and I spoke, CX meeting AI is the greatest transformational opportunity from a technology perspective, from a customer experience perspective. And the chance to build at scale at Cisco was an opportunity that was just impossible to pass up on. Yeah.

You're not going to get many like that. No. So you were the keynote this morning, day two of the event.

Most of the people watching this didn't have the pleasure of being here and seeing you. Explain to them, if you would, so that they don't feel FOMO, right? What'd you talk about?

All right. First, next time, please come in, watch it live. It is incredible.

We had a bit of a show. And let me tell you the heart of what we're trying to solve for. All of us are customers in our daily lives, so it's not a hard thing to fathom.

And often organizations are hard and complex to navigate. I always joke about, and I said that on stage, which is, if you stay on the phone long enough, you'll discover the company's entire org chart because you'd be bounced around, or you have to call a gazillion times to get something done. And that is what we call the coordination tax.

And all customers have to end up paying the coordination tax because systems are siloed. It's hard to work between them. 1% who's the uber, ultra platinum elite, you get a personal relationship manager, and they navigate the org for you.

The incredible announcement we made this morning was a piece of technology that will navigate the organization on your behalf, make you feel like the brand's only customer, almost get that brand to give a personal relationship manager for you at an economical enough number that it's scalable across the entire enterprise. So for you, you come in, no matter which avenue you approach the brand with, phone, message, physically, what have you, there's continuity of context, continuity of experience. There is an agentic harness that powers the AI agents, of course.

It preserves this continuity, keeps the context, runs agents long after the conversation's done to get you want to get done, and navigate and pay the coordination tax, so to speak, without you having to navigate it yourself. So this is the most ambitious piece of technology I personally, as a founder and an entrepreneur, have ever worked on, and I'm glad I'm building it at Cisco. I love it.

I love that. The day before, actually two days before, I went on a tour of the Cisco offices here in Austin, and it was a bit of a deja vu, but a revelation at the same time. I'm also a serial entrepreneur, started several venture-backed companies.

Back in 2001, late 2001, after 9/11, I started a company in Boulder, Colorado. Uh-huh. Mobius Venture Capital was the VC, and we were in their incubator.

Back then, incubators were all the rage, right? Yeah. And so just when we moved in there, they were shutting the incubator down, and us and another portfolio company took over this space.

Other than it was very Colorado with fieldstone and wood kind of thing, functionality wise, layout wise, from the phone booth areas to the smaller conference rooms and the touch pads and then the larger areas, the long tables where our developers would sit with the lights out, always the lights out because developers don't like to be in the bright light. It was very similar 25 years ago. What was not similar was the Webex, the screens, the functionality that came through those screens.

Many people, I think, still think of Webex as they're comparing it to Zoom. That's not the game here, right? If I had to emphasize one thing to the audience is you need to look at Webex not as your video conferencing solution, but exactly what Vinod, what you're talking about at your keynote today, right?

What does it bring? There's so much richness and goodness that can be garnered from using this technology like this. What do you think is the, not the barrier, and you have to know this, right?

Because your salespeople have to know this. What is the magic thing, if you will, that makes a customer say, "This is a must-have. I really must have this"?

Since you started with the workplace, I'll start there. Yeah. Then I'll come to customer experience.

I say this, obviously, I don't run that business, but I say this as a consumer of that experience, the workplace experience. In person is probably the most valuable mode of work because what you do- Oh, you're right ... in person is unsurpassable.

At the same time, we do have distributed teams. " The speaker is able to channel that and act as a funnel and make sure everyone in that real estate has an equal voice, equal listening, so to speak. It does the same with the way the cameras are set up, dynamic tracking, and everything for remote attendees, too.

But the most important thing is the way you manage this estate is with AI. So what a room where you have to say, "Hang on, this is my video. How's the visibility of the thing?

Is the audio good? " There are just so many things. And if you remember the sound systems of old, someone will come home, set up a sound system, and do this whole acoustic test.

Right. And it's this fine-tuned engineered stuff that only the highest end of sound system designers would do. Sonos.

Exactly. " And maybe it does, maybe it doesn't. They're bringing that enterprise-grade, AI-driven ability to build, operate, manage, optimize to the entire workspace, not just a room, but to the entire facility.

That's what integrated workplace experience was. Everything from the switching, the connectivity to spaces, the navigation, the manageability, optimization, humidity, temperature control, and ultimately, the audiovisual experience, all controlled, managed, built, operated from the same place. And the reason I'm saying that is all I know is the end experience- Right ...

which is superlative, right? So that's the workplace experience. " It's just that delivering N equal to one personalization and vital of experience is insanely expensive, very time-consuming, very labor-intensive, very process-intensive, just very expensive, and it's just not affordable to do it.

So the big thing that we are trying to solve for is, since you have the cultural bent to deliver great customer experience, can I enable technology for you to bridge that gap and be able to do that for pennies to the dollar? Which means every customer feels heard, respected, valued, and serviced the right way, but you do that in a profitable way. We need to bring these two together.

Right. It's not a mutually exclusive one wherein only the customers can be happy, or we can make money. Both are possible, and that's really the big pain we're trying to solve with the CX landscape.

I love it. And I agree with you. I think, look, IT has gone through its own journey of becoming a sinkhole versus a cost generator versus a profit generator.

Yeah. I think we have to have a similar view for the web CX stuff. Right?

What am I enabling- Mm-hmm ... by having this positive experience, whether it's customer to salesperson- Sure ... colleague to colleague.

Either way, when we work better together- Mm-hmm ... better things happen. We're more profitable, we're more productive, we're better.

And that, if you look at that strictly as a cost center versus a revenue-generated center, a revenue efficiency center, let's call it, I think you're looking at it all wrong. Right? And I think that's part of what your keynote is about.

Part of what you're trying to get across here is, guys, this is a game changer. Mm-hmm. Right?

This isn't certainly not what I had. And by the way, we had Cisco equipment back in 2001. Oh, sure.

It wasn't this Cisco equipment, though. And so this is a game-changing kind of experience. I agree, of course.

So to me, I think the big elevation that has happened in the game is ultimately, there's an old cliché, which is everyone is in customer experience. Right. Everyone's in customer experience.

But our technology systems and processes don't back that sort of logo, if I may say, but that's what we're changing. We have the very enviable position of owning a UCAS, a CCAS, and a CPAS platform, all operated, managed from cloud control, one single pane of glass, manageability, visibility, data synchronization, what have you. And so the vision really is to build one single holistic customer engagement platform.

And in the service of customers, that might happen in your physical point of presence, store, branch, pick your poison, at your digital storefront. Right? It could happen with you, or it could happen with your personal AI agent.

It may involve the person in the frontline, could involve someone in the villages of the organization. It could be in the contact center. Sometimes it might be a support case, sometimes it's a sales case.

We have siloed the org and said, "Hang on, this is about shipping. Go to shipping. This is about returns.

Go to... This is about marketing. " AI has this incredible ability to be the layer on top that navigates this organizational complexity.

You actually get a single window into the brand. Let's take the phone analogy. Okay?

As I said, it could be personal AI agents, it could be digital, it could be phone, whatever. But let's take the phone because the easiest to digest. Whether you call my branch or whether you find my phone number at my desk on page 17 of my website somehow, or you come to the contact center, or your call hops between the two, or you want to apply AI, you should have one platform to build, operate, manage all of this.

" "Fantastic. " "You know what? " A simple phone call that would've taken a store staff that is probably helping you buy a cooking oven or whatever, right?

Mm-hmm. Was then reduced with AI for pennies to the dollar. You service the customer, no phone went unanswered.

Store staff are still engaged with you, with the physical customer. Sure. And you actually did a $50 order.

Customer's happy, you made money off of it, pennies to the dollar. That's the kind of experience we need to give. But the incredible part, if I may just finish the thought, is whether that happened or the call got sent to the call center or whatever, context preservation is key.

Which means it doesn't matter if Vinod picked up the phone, the AI agent did, or Bob did, or Jim from finance. It's almost like you're continuing a conversation with your favorite human. There's continuity of context.

There's this persistence of knowledge, and it's still like, "Hey, I just sent you a message. " Uh-huh. That's what I think Jitu mentioned, Anurag mentioned- Right ...

which is give every customer a personal concierge kind of agent. Yeah. That's the vital of customer service we're aiming to offer to all of our customers.

What a great vision. What a great-- Because my next thing I was going to ask you is what about AI and agents? So you led us right down there.

I think, and I'm talking now as a personal consumer- Sure ... right? " Because if I'm calling about something that's in their bucket, it's a great experience, right?

You call the airline, I want to change my reservation. "Okay. What's your confirmation number?

" But I call the airline, and I've got a problem, not select one, select two, select three. And, "Okay. " And you say it, and even if it's an AI, it doesn't help, right?

What you are talking about, though, is a higher level of agent. Mm-hmm. I'm trying to think.

A better understanding. I'm going to go deep and geek out about this. You want?

Go ahead. I can hear it. I'm doing this.

I love what you're saying. So, let's start with just how AI agents have progressed, okay? Good.

The GPT moment happened approximately four years ago. Right? November 30th, 2022.

So since then, AI has gone from being a toy. We used to write poems. If you remember, the first wave was write me a poem about my high school reunion, and kaboom.

Oh, wow, it's amazing. Absolutely. So it was a toy, fantastic toy.

It became a tool. Tool is the chatbot era, which is I have a question, here's the answer. You go to knowledge base, run RAG.

" Right? Then it also did a couple of simple things. " You can check the knowledge database for thing and say- The calendar and do- ...

noon is open. Do you want noon tomorrow? And that is great, but they were constrained by one simple thing.

The construct of the AI agent was either conversation aware or the sophisticated agents were journey aware, which is, hey, I want to rebook my flight. Rebooking, option, refunded. That's fairly whatever.

" So you, let's say refund. "No, I don't want a refund. " Now it's like, hang on a second.

I know refund. I can go to the knowledge base and get you the refund information, or I can get you this other ticket information. Like either or.

I can do both. Okay? " It's like, yeah, but okay, that's an entirely different conversation.

Okay? I don't have a knowledge base. If it were a human with that AI agent's knowledge base, this is the answer the human would give.

The AI doesn't know that. So AI's like, "Hang on. " Right.

One or the other. And soon you're screaming representative. What we introduced was very, very significant in that direction.

We announced Dialog, which is basically the agentic harness that sits beneath our AI agent platform, and it has four superpowers. The first is the ability to run long horizon task agents. Agents that start in the conversation or after the conversation, run for days, weeks, and months, as opposed to be conversational obsessed.

They have a second very important, it has a context engine that takes enterprise context, which means you upload enterprise policies, documents, ethics guidelines, refund rules, brand guidelines, okay? Everything. Anything.

Exactly. And you upload prior conversations. This is what a good refund experience looks like.

This is what a bad refund experience looks like. It learns like a digital coworker. So AI going from a toy, to a tool, to now a digital coworker.

Digital coworker. So you don't build this AI agent. You train, onboard, optimize, and continue to up-level this agent like you do a digital teammate.

So it takes that, it takes customer context. Who are you? What's your history?

What's your whatever? What's your satisfaction with it? And it creates something called the relationship context.

This is N equal to one personalization. 2 million miles, you had a flight delayed two weeks ago, unrelated to this conversation. Okay?

So the example we showed on stage was you asking for a refund. Theoretically, the 24-hour window has passed, but there are many other contexts that are relevant to this. This journey was all around different.

Simple answer was not a lot. " Now it says, "Hey, hang on. I know you've been platinum.

I know you've been with us for 10 years. 2 million miles with us. I also know that we delayed your flight two weeks ago.

So hang on, let me put this together. Let me go to the human, because AI can't exercise discretion- Discretion. Right ...

" Discretion. Right. So it takes all these random things, puts it together, navigates the org, the human, the policy, works with your personal AI agent onto your travel insurance- Right ...

and it gets the job done. So the biggest thing with the agentic HANS that we built underneath, which is dialogue, is it is now relationship aware, like a personal relationship manager is, beyond being just conversation or journey aware. And that elevation conversation to journey to relationship, and armed with the context engine that it has in the back, allows the agent to have superpowers, which it just never had before.

And the last thing, it learns per conversation. Here we go. It gets better, like a digital coworker does.

We're all better at our jobs, I'd like to think so, than we were on day one. And why is AI not? Because if you run a simple AI agent, knowledge base- It never progresses ...

right? Same question, same answer. Which is fair.

That's the whole point, right? You've told it to listen. These are your guardrails.

You're this agent. You're going to listen to knowledge base. So that's what we are transforming.

So the underlying HANS, a technology that didn't exist till not too long ago, powering the agent, now allows it to do the things you're asking it to do that today's circle of agents just can't do. So I'm genuinely excited to bring this exact technology- I'm excited ... to our customers.

What an exciting world. It's an exciting world indeed. When is this available?

So we currently have design partners. You heard from one of our design customers who was on stage with us. So they spoke about what problems they're solving with it.

So the way all software is being built now is you don't just build and they will follow. You build with and for customers. So we have a set of design customers building it with us.

We expect beta to be Q1 next year, but the door to design partners is open. As all products with AI, it'll be hyper-optimized to industries and contexts and journeys. Right?

And as more design partners come in, we'll expand the aperture, sort of open up the aperture a bit if you see. But Q1 is when we are announcing beta, but the design partnerships are open. Work is happening on it now.

So if you have a use case that needs these superpowers, we are open to talk to you. I love it. Now I know why you're in the position you are.

Thank you. Really, that's a vision. Thank you.

That's a vision. It's very clear. And the vision always wins.

In Jeetu Patel's product organization at Cisco, all vision better be reality in 90 days or you're not building it. That's the way that is. No, thank you so much.

That was fantastic. Hey, we've got a lot more coverage here from Webex One. Stay tuned.

You're watching Techstrong TV.