How Agentic AI Will Reshape Customer Experience | Rangan Gangavaram (Verizon)

**Les** (0:07)
Data Storytellers, today I am here sitting with Rangan Gangavaram from Verizon. And Rangan and I have been having conversations, I think for the past two years, one and a half years or so, and then we reconnected recently, had some great discussions on what's happening out there today in the field of AI, especially how that affects data and analytics as a function, as a profession. And I just wanted to immortalize some of those insights and share it with the community. So Rangan, first of all, welcome to the show.

**Rangan Gangavaram** (0:41)
No, thank you, Les. Thanks. And I think I distinctly remember our conversations about a year and a half ago, and I was very keen on some of the partnerships with you, and I'm so glad that this happened right now.

**Les** (0:57)
Absolutely. And there's a lot to talk about, and probably there's more to talk about than last year. I think we connected last year before our Nashville session that we had, and that already had some AI themes, but since then, things really accelerated. So, Gen.AI was the spark that lit everything on fire, and now AI is the talk of the town, and what we're seeing is that all these narratives are floating around in organizations, and it's not easy to figure out what are the best bets and what are the best moves. So, first of all, just to kind of cut into one of the hot topics. Agentic AI, right? So, let's start with a broad question. Agentic AI, do you think it is the real deal? Is that certainly the direction where we're headed? If we look into the future five years from now, are agents all over the place creating all this value and doing work that humans used to, or is it just kind of a fluke, and another trend will take over soon?

**Rangan Gangavaram** (2:00)
Yeah, I think it's a real deal, no doubt about it. And if we think about the reasons why it's a real deal. So obviously, if you look at the history of automation, so RPA comes to everybody's mind. And when RPA was exactly in the transformative phase that agentic AI is today, the questions were very similar in nature, which is, well, what can RPA do? Well, can it scan through documents and find out where should my contract negotiation be focused on? It could. Can it parse out things which would take a few FTEs for X number of days to decipher what's in a code? It could do it. So you think about RPA as a tool, it is real, it has added efficiency, it has added productivity to people. And we are at a place where today we can clearly see that without automation, I wouldn't have really thought about how I could have expedited the journey. Agentic AI is in the similar path. Now, what's going to be different, right, is that agentic AI is going to be a plus plus on top of RPA, right? And this is where the ability to reason, right, and ability to take action on behalf of a human, right?
And those are areas where we've seen this play out time and again, right? You pick up a phone, you call a customer care, and you say, hey, by the way, there was an incorrect billing, right?
And the agent should know that, hey, by the way, a week ago we had some sort of an outage which impacted 7 million customers in their billing, right? So when I call them, it's not news to them, right? So which basically means that they'll be like, I'm sorry, you had a billing issue. Here's what I'm going to do. I'm going to revert the bill to your original bill, and I'm going to do that for you, right? Now think about the same scenario play out in the agentic AI, right? It knows exactly that I was part of a billing issue which happened a week ago, right? And now the agent knows exactly what was the billing issue, and it also knows the dollar amount that was impacted. So think about an agent acting on behalf of a car care center supervisor to say, this is a legitimate caller who has a billing dispute. It's a no known, right? And I can reason out because we just dropped the bill yesterday, and he would have responded to it, and that's the call. So how about taking an action which basically could lead to even a surprise and delight, right? Which says, hey, by the way, I know, are you calling on behalf of a billing dispute that you uncovered yesterday? And if so, we are happy to tell you that we're actually waiving that $7 off of your bill, right? So it's going to reason out, it's going to make decisions, and more often than not, it's going to do it at a pace, at a way that it's faster than what would have happened, right? So that's an example of where we're actually seeing initial success of agentic AI, where we actually look at our calls, look at our call transcripts, understand the velocity and the momentum of the reasons why it's bubbling up, and if we can identify the people who are actually part of that, we can actually take more proactive and preventive actions. And that's a reality today. It's in a POC. And I only see that kind of use case as being exploded over the next couple of years, where you can reduce call volumes, you make better decisions, it's improving the CX, right? And it's also hitting the problem at its nail, which is, do you know the problem a customer is facing even before they call? And the answer is yes.