AI Implementation and ROI – Techstrong AI Podcast EP62

Hello and welcome to the Techstrong AI Podcast. I'm Amanda Ani, and with me today I have Wendy Collins. She is the Chief AI officer of NTT Data.

How are you doing today? Hi, I am doing really well. Thank you, Amanda.

Happy to have you on the show. Can you share a little bit about NTT data and what do y'all do? Sure.

NTT data is a global technology and services firm. Um, we are headquartered at the global level in Japan, um, and I am in the North America business entity, and we bring the best of technology, including data and AI to our clients across the globe and to our clients here in North America. Wonderful.

Well, we just had the NTT upgrade conference just a few weeks ago. Lots of great information shared there, especially in regard to ai, which is a big topic. So, um, let's start on, um, what is, uh, from a broad perspective, what are some of the concerns that companies have when it comes to implementing AI's technology?

Yeah, for sure. Well, um, thank you for mentioning upgrade. It's one of my favorite events that NTT sponsors every year, and it really highlights all of the forward thinking research work that we're doing both with, um, academia and other organizations.

And what upgrade really does beautifully is it stitches together the research work with what's really happening on the ground with companies and clients all across the globe. And, you know, you, you can't throw a rock very far without hitting a conversation about ai. And you're absolutely right.

One of the questions that a lot of, um, enterprise business executives struggle with is, okay, how do we capture the benefit in our p and l in our competitive landscape of ai? And I would say, um, one of the challenges is that clients often start with, um, AI implementations that impact operational efficiency. And that's a great place to start.

Don't get me wrong, especially if you're very early in your AI journey, that's a great place to dip your toe in the water. But to be quite frank, it's very hard to actually get measurable benefit in dollars and cents from operational efficiency. Um, and so what we really encourage clients to do is to think bigger in terms of capturing value and value creation by focusing on use cases that are, that are not just operational efficiency, but can unlock net new revenue streams or that can help you take advantage, uh, of your existing lines of business and drive, um, deeper revenue retention there.

Uh, the next level up in terms of value capture and value creation is really around, um, customer engagement or in the case of healthcare patient outcomes. What are you doing to retain customers to prevent them from churning, from acquiring new customers? How do you make it faster, easier, more compelling for customers to come and want to do business with you?

And then the, the top tier in terms of, uh, use cases for AI that we really encourage clients to focus on is how do you gain competitive advantage? How do you leapfrog your competitors? Or if you are the incumbent in terms of your, uh, competitive landscape, how do you strengthen that moat around your competitive advantage by leaning into ai?

So the challenge is, um, to move up the value creation ladder and not just focus on operational efficiency. So then when companies do decide to try to, uh, implement a solution for a certain problem that involves AI technology, from your experience, where are some of the roadblocks on this change journey? Yeah, um, well, so the first one is not recognizing that just because you have the data that you think you need for a particular AI use case, it doesn't mean that it's AI ready data, right?

Um, so this is a conversation I have a lot with, uh, executives who they say, Hey, why can't we just launch into starting this AI engagement? We have the data. And so part of what we have to talk about is what it means to have AI ready data.

Um, and oftentimes, um, what that means is moving it out of your operational systems, the the systems that run your business every day and move it into an analytics environment and stitch it together in a way that AI models, AI solutions and AI analysis can take advantage of. Um, so I would say that's one of the roadblocks. Another roadblock is when your AI use case is focused, um, really and limited in, in, um, in its, I'll call it buy-in to the technology team, right?

And, um, one of the real hallmarks of success for AI projects is bringing in the people, the individuals oftentimes frontline, uh, individuals in a manufacturing plant, it may be or frontline workers in, if you have a, a back office, um, AI solution, let's say for, um, for your legal department, it may be bringing in the actual attorneys that are working on briefings and working on contracts, bringing them in at the very beginning and making sure that you are thinking through the business problem with them so that when you design your AI solution, it's really getting at the heart of the problem that your end users care about. So that's number one. And just as important, it's also about, um, bringing them in early to drive adoption.

You know, one of the things that I have, um, discovered as we start really leaning into generative AI and ag agentic solutions is that people who have the exact same experience, background, the exact same education, the exact same lifestyle, may have very different comfort levels with using generative AI and AI agents. And so identifying what that level of comfort is and leaning into embracing people where they are and bringing them to where you need them to be, um, is a really important step in overcoming that roadblock. Yeah.

So communication is really key. That's exactly right. That's exactly right.

Alright, so let's talk about, um, there was a, um, a big discussion around a new group, uh, the physics of artificial intelligence group. And I would love to hear your thoughts about the importance of this group, um, what you, how you think it will impact the future, and, um, maybe just a little bit about, um, what problems it's, it's hoping to solve. Yeah, for sure.

So what I love about this new group is they are tackling one of the hardest problems in ai and that is understanding why generative AI models are doing what they do. Why are they making the recommendation? I mean, at their, at the heart of all generative ai, it's it's math, it's largely math.

Um, but they have gotten so complex in terms of their ability to, um, in some cases reason, uh, and come up with answers and ideas. Um, but we don't have a really solid understanding at the individual problem and question answer of why they're getting to those answers and, uh, responses. And so this team has taken on, um, a really, I think, important challenge.

And when we talk about the physics of ai, what we're really trying to understand is what are all of the steps along the way that get an AI solution to the answer. You know, we, we often use this phrase ai or the the acronym ai, um, to, to mean specifically generative AI and a GenX, um, form of the, um, AI continuum. But the, the AI continuum actually is quite broad.

Uh, and so I often use AI to mean the umbrella term that is AI that starts all the way over here on the left hand side as um, you know, business analytics and insights and descriptive analytics that we've been doing for a long time. And the next step in the continuum is, uh, data science. So think predictive and prescriptive modeling and we have machine learning and as you keep going towards generative ai, the transparency of our models get less and less.

We understand less and less about how the model actually comes up with the answer that it comes up with. So the physics of AI team is really tackling as you move to the right on that continuum, how are we going to be able to understand the why? 'cause that's where the trust comes from.

That's where you get adoption. Going back to the, uh, conversation we just had moments ago, right? Is when users understand the why, that's when they're willing to accept the what.

Absolutely. It sounds like very important research being done in that department. So AI is advancing extremely rapidly.

So what is your advice for how companies can stay ahead of it and get full ad advantages from AI tools? Yeah, so well, thank you for asking that question. I think it's a really important one, and part of it is an honest understanding of where you are in your AI journey.

You know, here at NTT data, we work with clients that are at all different points in their AI journey. We have some clients that are very early, um, they maybe have never even taken advantage of some of those left hand side enterprise AI continuum tools like data science and descriptive analytics. Um, and then we have some, some, uh, clients and executive teams that have already started making their way into, uh, the far right hand side of the enterprise AI continuum.

So, uh, an honest assessment of where you are and an honest assessment of what your internal capabilities are. Um, so, you know, we see clients who, um, they are well set up for success because they have been investing in tools and teams for a long time. Um, and then they just look to, uh, partners like NTT data, for example, for, um, thought leadership and understanding of where other companies like them are taking advantage of ai.

Um, and then there are clients who they need help setting up an AI strategy to really get their AI journey off the ground. I would say the most mature clients understand that they need some form of governance around their AI process, tools and capabilities. Um, and that's actually what, uh, one of our clients that was on a panel at Upgrade was sharing is how, uh, they at Clario work through how they decide what are the right AI projects to go after, how they decide what are the right AI tools to bring into the environment and how they decide how to early in the process design and, um, determine how they're going to measure success once it goes live.

Wonderful. Well, if there was one key takeaway you could leave our audience with today, what would that be? Um, the key takeaway would be, if you haven't already started, get started, this technical landscape that is AI is moving faster than any, uh, technology breakthrough we've seen in history.

It's moving faster than cloud, it's moving faster than the pc, it's moving faster than internet. It is moving so fast that if you don't get started, you will find yourself left behind. So please don't wait.

Thank you so much for coming on the show and sharing your insights with us today, and I agree it's it's the biggest technology I've heard about in a while. Yep, exactly. Well, thanks so much, Matt.

It's been really fun talking to you. All right. And thanks to our audience as well.

Stay tuned. There's more.