AI-Empowered Customer Service, From Hype to Scalable Operations - with Shri Nandan of Comcast

**Daniel Faggella** (0:12)
Welcome, everyone, to the Emerge AI in Business Podcast. Today's guest is Shri Nandan, VP of AI Products and Experiences at Comcast. Shri examines why AI deployments in enterprise customer service so frequently fail to deliver business value. The conversation addresses how context-rich agent existence change what resolution means in practice, why human agents remain essential for emotionally complex interactions, and how a conservative stage draw out reduces the risk of large scale failure. Today's episode is sponsored by Dialpad. A quick note for our audience that the views and opinions expressed by Shri on today's program are her own and do not reflect those of Comcast or its leadership. In this episode, we cover how to move from AI proof of concepts in customer service to deployments that consistently improve business outcomes. To go deep on this topic and learn how consultants are winning business with evidence-based AI ROI and building long-term capabilities instead of chasing short-term gains, download our free PDF report, 3 Keys to Thriving in the Coming Era of Automation, at emerge.com/cok1.
That's emerj.com/cok1 to download your copy. Now the conversation with Shri.
Shri, welcome back to our podcast.

**Shri Nandan** (1:46)
Thank you so much for having me.

**Daniel Faggella** (1:47)
Today, we're basically building on our last episode, where I want to start with something that I think has been at the top of the mind of many of the leaders that we talk to. There's this enormous amount of investments flowing into AI and customer services right now.
But yet it feels like we have the gap between what gets deployed and what actually delivers business value. And that gap is still striking at this point.
So I'm wondering for you, having driven this inside financial services, healthcare, and now at the scale of your current organization, what has historically stood between the ambition and the business result?

**Shri Nandan** (2:25)
I think the question always remains, are you solving a problem? Whether it's healthcare or financial services or telecom or anywhere else, or it doesn't matter what the technology is, if you are not solving a customer problem, then it doesn't really matter and none of it matters. If you don't solve customer problems, you are not going to generate revenue, you're going to lose customers. So that is the first question everyone needs to ask. When you ask that question, it opens up the ability for the company, for an organization to say, these are the lowest hanging fruit and the highest impact items that I want to go after based on data.
This is what the customers are telling me, and that's what I'm going to use the technology for, to have maximum impact, and then I hope to move the needle in my metrics in a certain way because I've used the technology, and that is all you need to do. When you start using AI for the sake of AI, or when you start using technology because it's cool, or because it's fancy, and building proof of concepts that don't really make any sense and don't really solve a customer problem, then nothing is going to happen. It's just money down the drain, and you're not really going to move the needle or show any positive results in your revenue in any sort of way. So I think really the basic question that everybody needs to ask is, why am I doing this?

**Daniel Faggella** (3:51)
And that makes sense. And I think we have those listeners that they hear you, they understand what you're saying, and they're maybe about six months into deploying their CX AI, and they're thinking, okay, how do I know if this is actually working? So I'm thinking, what is a failed AI deployment and customer service actually look like in six months in? Like, can you walk us through what the team will be experiencing? What is the real feelings? What is the real red flag that they'll pick up on?

**Shri Nandan** (4:18)
Some of the things that you'll start to see right away is burnout. Because when you don't have all the foundations in place and you haven't really established a good north star for the team, what you're really building doesn't really have a final common purpose. So that's the first sign that you start building something without a purpose and you start doing more and more and more, and this team starts to burn out. The other thing that you might see as a big red flag is that your technology is getting more and more cumbersome and heavy.
For example, if you're building out an agentic framework and it's getting bigger and bigger and bigger, but it's also because you don't have time and you're constantly under pressure to go put things into production, you're building something that has become really riddled with tech debt. There are so many problems, there are so many corners you've had to cut because you have all this pressure coming at you to show something. So you're going to see a list of items that's growing that we call tech debt, and that's going to be the reason that your initiative will most likely fail. The other red flag that you probably want to keep an eye out for is, if you've been doing this for 6 months and nothing has changed, your CX metrics are exactly where they are or they're getting worse. If TNPS is bad, if the repeats are bad, if all of that is still the same or getting worse, then you need to stop and take stock of, do we really need to keep doing this? So team morale, KPIs, the integrity of your technology, these are all things that will start to show some clear signs of problems in the six months in if you did not start off on the right.