How Mastercard Stopped $13B in Fraud With Real-Time AI

0:00The line between product and technology is becoming more and more fuzzy, right? You think about how companies are developing, you know, product organizations are becoming technology organizations. It's becoming much more gray.

0:13Peter High: Welcome to Technovation. I'm Peter High, president of Metastrategy. My guest today is Nirav Mehta. Nirav is the chief technology officer of services at MasterCard. MasterCard is a global technology company that powers commerce in more than 200 countries and territories. The company generated nearly$33 billion in revenue in 2025. While MasterCard is best known for the card in a consumer's wallet, its role extends well beyond processing payments. The company provides services that help customers secure transactions, prevent fraud, personalized experiences, and turn data into actionable intelligence.

0:46Peter High: Nerov leads the engineering and technology organization across the fastest growing part of the company, MasterCard Services. His responsibilities include data infrastructure, analytics platforms, core technology capabilities, and the team supporting the company's services products. Before joining MasterCard, Nirov was Group Chief Technology Officer and Chief Information Officer for Corporate Engineering at the London Stock Exchange Group. He brings more than 25 years of technology experience across financial services, healthcare, education, media, and telecommunications. In this conversation, he and I will explore the parts of MasterCard that consumers may not see, including the services surrounding each transaction.

1:25Peter High: We'll discuss MasterCard's data commercialization platform, the principles and governance required to use data responsibly, and how artificial intelligence is helping the company turn data into intelligence and intelligence into action. Nirav, welcome to Technovation. It's great to see you today. Thank you so much for having me. I'm excited to be here. Today's episode is brought to you by Tynes 3B. Organizations are pushing teams to move faster with AI. That's creating a wave of AI-generated code and unsanctioned automations, leading to rising costs, security risks, and huge IT burdens. Meet Tynes 3B, a single, secure environment that enables employees to build with AI while giving IT and security complete control to run and govern that work.

2:05Peter High: 3B enables effortless building. Describe what you need in plain English, and 3B builds a fully functioning workflow. It uses IT-authorized connectors, meaning credentials, state-protected, and non-technical teams can innovate without taking on unseen risk. Also, with full visibility across every workflow, IT gets total governance, while features like Auto-Tune and Auto-Fix ensure workflows improve and self-heal continuously over time. Employees get the freedom to build, while IT gets the control to govern. No shadow AI, no data leaks. Tynes 3B moves IT and security from blockers to enablers, giving teams the speed to innovate without compromising control.

2:43Peter High: Visit the Tines website at tines.com. That's T-I-N-E-S.com to learn more and try 3B for yourself. And now on to the interview. Likewise. Well, I don't think your company MasterCard needs all that much of an introduction, but maybe just because there's the aspect of the business that I think is most commonly understood and a lot of it that is not, maybe you can provide a little bit of an overview about MasterCard, and then I'd love to further get into MasterCard services, the part of the organization where you spend most of your time. Provide a little bit of background, if you will. Sure. First of all, Peter, thank you again so much for the opportunity today.

3:19So, you know, everyone thinks of MasterCard traditionally as a payments company, but the way we think about ourselves is we're really a global technology company in the payments industry. You know, our mission statement is to really connect people, businesses, institutions, merchants, governments, to really drive commerce at scale. So really more than a payments network from that perspective. Beyond processing, we really help businesses and customers with complex business problems. One of the little secrets is we are one of the largest cybersecurity companies. So we do a lot of work around protecting transactions, providing cybersecurity solutions.

3:53We provide lots of insights and analytics to enable commerce. And now with the advent of Agenda Commerce, we're playing a huge role in Agenda Commerce with enabling machine-to-machine payments. and also enabling other types of payments in the agentic space from that perspective. So that's a little bit just about us from a high-level perspective. I'll talk a little bit about my rebate, Peter, if that's helpful.

4:19Peter High: Please do. Yeah, I'd love to have you talk about that. And in the process, talk a bit more about the part of the organization that specifically that you're a part of. But thank you for that overview. Yeah, so if you think about MasterCard, we obviously operate the payment space from a technology perspective. But across the payments lifecycle, we provide a variety of value-added services. I touched on earlier, everything from cybersecurity solutions to personalization, e-commerce, marketing services. We provide business insights and analytics. We also enable financial data for consumers that are available from a financial data perspective.

5:00And so those services I mentioned make up the services business of MasterCard. The services business of MasterCard is about a$14 billion business growing about 20 % year on year, makes about 40 % of revenue of MasterCard. And I have the privilege of leading the engineering and technology organization for the services business. So my team is responsible for the product engineering across those four product segments I mentioned from an engineering perspective. I'm also privileged to lead all of our analytics, all of our core data platforms, and all of our decision capabilities across the business.

5:36So it's a broad remit, about 5 ,000 folks across 30-plus different locations operating at a global scale.

5:46Peter High: Incredible. And it's really, I mean, it's remarkable. You and I have talked about this previously off the record. just the extent to which the breadth of your business, the number of transactions that you're involved in, your ability to see the trend data associated with payments has really profound implications and possibilities for insights as to economic activity in different parts of the world and how that's trending, among many other things that one can do. I can only imagine it must be a fascinating exercise to think about how you leverage the remarkable data you have at your disposal to create these services that you've just described.

6:28Yeah, you know, it's interesting, you know, Christmas time, you know, people always see, you know, what's the hottest toy or what are the travel trends, you know, that, you know, happen in the marketplace. We actually have a Mass Credit Economics Institute. and one of my colleagues, Michelle Meyer, who's our chief economist, you'll see her on CNBC, you'll see her on CNN, you'll see her on various outlets. And we're always talking about kind of world trends, you know, from a macroeconomic perspective. You know, what are we seeing from a geopolitical perspective, but also from just a commerce perspective?

7:01You know, what are travel trends, you know, from a summer, from a fall perspective, where are consumers spending dollars, et cetera. So we really have a really good view of what's happening in the economy from that perspectives.

7:13Peter High: Can you provide a bit of background into the data commercialization platform that your teams help to drive? Yeah. So one of the largest, one of the remits I took on when I came to this organization was really looking to modernize our data infrastructure at MasterCard. You're looking at how do we build a cohesive kind of enterprise data foundation to power not just services, but the entire organization. So our core business, consumer payments business, our commercial business, and our services business. And the goal here was to build a platform, not necessarily a technology platform, but a platform that really drives innovation, decision-making, really ultimately driving business outcomes.

7:55And the goal here was to take – we have lots of data at MasterCard, some that we generate through our core processing, some that we get through partners, some that we acquire, et cetera, And all that data prior to the DCV platform, data commissioners and platform, was sitting in a variety of different silos. And so we've built over the last several years this new platform that really brings all data together. And the goal of this platform is to really help drive consumer insights and build new products for our customers, doing it in a safe, trustworthy perspective. Right. Having governance, resiliency, security, privacy.

8:38Those are all key tenants of the platform. Ensuring that we're doing everything. You know, the consumers ultimately control their own data sets. But, you know, providing a platform that covers all those principles, all those tenants from that perspective. And this new platform, you know, from a technology perspective, it's multi-engine. So we're using working multiple partners. Interopoli is really critical. You know, there's the common technology term iceberg. We're using an iceberg kind of format, which can be leveraged across different cloud providers, as was different technology stacks. It avoids long-term technology lock-in and really provides a unified governance.

9:15Again, maintaining MasterCard standards around trust and compliance from a safety and security perspective. So, again, that's been a really critical piece, and it's really the kind of bedrock of our data foundation across the enterprise. And on top of that, foundations, all our products and services live on top of that.

9:31Peter High: That just sounds remarkably complex, Nirav. I can only imagine it's easier said than done. As you think about bringing to life what you've just described, at least at a high level, can you describe sort of the process of doing that? Again, I can't imagine it was a light switch where you turn this on and everything came together, but rather something where you had to have a meaningful plan in place with a hierarchy and layers associated with it. Can you describe that? Yeah, look, we started with, you know, Peter, like a simple use case, right? Looking at, you know, I'll give you a great example.

10:07we have a lot of data around location, right? Which is used to power a variety of products that are location aware, you know, from a geography perspective, from a zip code perspective, from an address perspective, et cetera. And, you know, that data you set, you know, across multiple platforms and that data would become very stale very quickly, right? You can think about, you know, addresses, people moving, you know, businesses relocating, et cetera, you know, and data would become very stale. keeping that data updated is a constant kind of challenge. And so that was the problem statement. How do we provide this data in a more meaningful real-time fashion to really drive a variety of use cases around commerce?

10:49And so we took that use case, built this platform from a DCP perspective, and really used DCP to deliver that outcome. The goal here was to take data that would come stale over 8 to 10 weeks to deliver it within hours. So taking a process that would run multiple weeks and getting it down to probably less than 8 to 10 minutes in some cases in terms of performance and delivery to the customer. And so taking that problem, the goal was to bring the disparate data sets together, do data cleaning, cleansing, and then making it accessible for the overlying applications and services that support the data and making it available for consumption.

11:30And so that's just one example of kind of how we did it. All of everything we do, Peter, is really use case first, I would say, right? Engineering is really, really important, but it's really around what's the value we're driving, right? You know, everything we do is focused on value. What's the value to our customer? What's the value to our consumers? What's the value to our partners? And so that's how we think about building this platform.

11:54Peter High: I know from our past conversations and past conversations with colleagues of yours, the role that trust plays within your organization, a key element to making sure that MasterCard is an effective business. Can you talk a bit about the key components to reinforce trust with customers and the role you feel that you and your team can play to help there? First of all, I think trust is the bedrock of everything we operate. In our model, the consumer controls manage their data, but that's principle number one. If you don't have trust, you can't operate in this ecosystem. And so really security, privacy, trust, reliability, regulatory requirements, bias, fairness, you know, customer trust are all critical elements, right?

12:38They're the key ingredients of a recipe to really operate this ecosystem. And, you know, really, from our perspective, it's understanding the risk, you know, the conditions for experimentation, you know, from a trust perspective, right? When we create trust, we have lots of governance that's built in from an automated perspective, how data is being leveraged, making sure the data is resilient, secure, making sure customers understand how data is being used. And that's how we operate. A great example is we do a lot of work around fraud. And fraud traditionally doesn't happen just when the payment transaction is happening.

13:14It's happening much prior to the transaction happening where malicious actors are traditionally stealing credentials or doing code injection or skimming payments from websites or devices. And so a lot of our services around trust help support and prevent this. One of the great examples is we have a product within services called Mass Car Threat Intelligence. And this product really connects the dots between what we're seeing from a security perspective to what data is being exposed to the dark web to the broader ecosystem so that we can prevent fraud and really protect the customer and our partners and our businesses ultimately.

13:53So, again, trust is a kind of paramount thing we look at as we operate in the marketplace.

13:59Peter High: You mentioned the sanctity of use cases. You also describe the methods you and team are using to ensure your data is in a form where it can be leveraged for greater insights. Now, let's talk about how you take the foundation that you've built and the use cases you have now that particularly excite you around artificial intelligence. How are you layering that in and what are some of the powerful applications of that, if you can offer some stories there? Yeah, so look, one of the great examples, we recently won the CI 100 Award for, is called Decision Intelligence Pro. We call it DI Pro. And the analogy, the way DI Pro works today is, imagine you're at your favorite Starbucks in the morning.

14:42You're using your MasterCard to make a purchase. You tap your card against the terminal. When you tap that card on a real-time basis, we are looking at that transaction to determine if it's fraudulent or legitimate, right? Has your card been stolen? We're looking at billions of parameters that we look at. We look at purchasing behaviors. We look at fraud patterns. We're looking at a variety of data points. And this is all based on AI. We have AI models. We have hundreds of AI models, which really freed this intelligence. And through this, we drive graph-based intelligence around payment behaviors to prevent fraud.

15:24And so, between your TAP and the transaction being approved, we're applying AI models to look at that transaction and to prevent fraud. DI Pro today, just this year alone, has prevented about$13 billion of fraud annually, just so far to this year. And we haven't even reached the holiday season where obviously sales volume increases dramatically. And so that's a great example of kind of how we're doing real-time decisioning, applying AI to prevent fraud at scale.

15:57Peter High: And as you provide these insights to business customers, can you talk a bit about are there commonalities across the businesses you serve? Or is it literally around the globe and across all industries? Talk a bit about the sorts of companies you engage with, with the insights that you've been describing. Yeah, look, we have a variety of products that we market towards. Obviously, issuers, banks in the financial service space are key customers. But we also work very closely with merchants in the marketplace. But we also work with government institutions. You know, one of our other products I'll talk about from an AI perspective, Recorded Future.

16:40So Recorded Future was an acquisition that we did about two and a half years ago. And Recorded Future looks at threat intelligence data across the dark web. So imagine, you know, looking over a thousand different data points across dark web, not only just a financial fraud, but looking at threat intelligence around just geopolitical issues. cybercrime, terrorism, et cetera. And so Recorded Future aggregates all these data sets and makes these data sets available in a graph-based kind of platform capability for a variety of consumers, a large number being government institutions who are using this for a variety of use cases.

17:23Lots of law enforcement use cases from that perspective, looking at malicious acts, et cetera. And so we really operate across multiple, verticals, not just financial services, but governments, insurance, healthcare, et cetera, with that product set from the quarter future.

17:43Peter High: Amazing. Can you, aside from the sorts of resources that are part of the team as a result of the acquisition, can you talk about the evolving skill sets that you look for to populate your team in light of the work that you do? I can imagine as the state of the art relative to data, analytics, artificial intelligence, machine learning continue to advance, that the skill sets necessary, whether those are trained for with existing staff or recruited for, so that you're bringing people who know what good looks like in the areas that you're headed towards. Talk a bit about the evolution of the team and the skills that you look for.

18:18Yeah, you know, this is a great question, Peter. You know, I'll say one of the things that's kind of happening is when in engineering, kind of in technology in general, the advent of JetAI has changed the skill sets people require. Now, if you dream it, you can build it, if you think about it. You've got Claude, you've got Copilot, you can just imagine and develop it. And so, traditionally, you were looking for strong comp sci backgrounds around coding skills, etc. Those are important, don't get me wrong. But I think with the advent of JetAI, you're looking to ensure that the strong skills are on systems thinking.

18:56you know, looking at problem solving, right? Looking at, you know, looking at, you know, fundamentals of computer science, less more on the coding ability and more on kind of, you know, design thinking, right? Looking at how you design and build and deliver various products and services. Because, you know, one of the things, we had a great conversation actually with Microsoft about this earlier this year. One of the things you're seeing in the workforce, particularly with graduates coming out, is there is a cognitive decline happening potentially with folks relying on AI, right? Just purely for development of everything.

19:33And so it's a balance, right? You need to ensure that you understand how to apply the AI intelligently, how to apply what it's providing, understand the context you're providing it, understanding the response you're getting back from the AI perspective. And so, again, traditional skills are on ML and, you know, comp sci are critical, but it's really the critical thinking skills that are really important as well. And so that really drives a lot of thinking around that. The other thing I'll say is the line between product and technology is becoming more and more fuzzy, right? You think about how companies are developing, you know, product organizations are becoming technology organizations, right?

20:12The individual that's doing the research around product requirements to someone who's running the PRD, the product requirements document, to the engineer that's actually building the prototype, it's becoming much more gray. Because now you could have essentially the same person doing the requirements definition to doing the delivery. And for smaller programs, smaller use cases, we have much smaller pods. Right. Traditionally, you would have the pizza's buck sized engineering organization of, you know, five engineers, one UX person, one product person and one designer. And now you may have two engineers, you know, one product person and maybe 10 agents.

20:55Right. That are affecting the development. And so it's changing dramatically of how you operate. Right. There's a there's a fundamental rewiring of how you design, deliver, build and operate products and services in the marketplace. And so we're going through that cultural revolution, I would say, you know, from an organizational perspective. So sorry, long-winded answer, but it was a really rich question you provided. So I wanted to give you some context.

21:21Peter High: Not long-winded, but a rich answer, in fact. So thank you, Nirav. You alluded to earlier a topic that I wanted to raise with you again and dive a bit deeper into, and that is the evolution or introduction to a greater extent of agent-to-agent commerce and the extent to which we are, in some cases, already down a pathway towards that or perhaps on the cusp of something that's a bit more where this becomes a bit more de rigueur for people. And it's a different sort of commerce motion that happens there, needless to say, when it's agent to agent as opposed to individual to company or individual to individual.

21:56Peter High: And I wonder how, given the vast experience your organization has with payments, with commerce, how you think about the similarities and differences when you think of agent to agent commerce? Well, I think this is an emerging space, Peter. There are lots of perspectives. So I will share. This is my opinion. I want to put that out there. You know, there's a lot of emerging standards, right? So OpenAI has their standard for agente commerce. You know, Anthropic has their standard. Microsoft and Google have their own standards, right? Amazon has own standards. We as mass credit work with everybody, right, in terms of agente commerce.

22:37There's lots of experimentation happening at scale across the industry, you know, in terms of how do we make this real? What does it mean? You know, with agentic commerce, you know, is the agent operating on your behalf or is it purely agent to agent, right? Is it two machines talking to the other? Or is it Peter's agent that's operating based on a request you've made and then connecting and doing some commerce? You know, there's a number of variables around, you know, how do you verify the intent of the agent? How do you know the identity of the agent? Security, privacy, regulatory concerns all play significant concerns in that conversation, right?

23:13So this is a little bit of like, you know, the web, whatever web first came out. You know, a lot of emerging standards, a lot of good intent, right? Everyone wants to kind of be in this space, but, you know, it's really kind of an emerging space. And so like everyone else, I would say we are, you know, we're working with a lot of folks around this. You know, AI-enabled product innovation is happening across our entire portfolio. But, you know, it's, you know, there's some announcements coming out. I don't want to, you know, around Agenda Commerce and such from MasterCard shortly. So I would say stay tuned.

23:51You know, we made some announcements on our agent pay, et cetera, and there's more coming out. But really, the last thing I'll say on this topic is, look, when it comes to payments, you know, we have years of experience to your earlier point, right? We put a lot of trust and a lot of credibility in this, and we are also going to bring all that knowledge and all that experience to object to commerce from an experience perspective. So more to come, but that's kind of our point of view at this point in time.

24:19Peter High: Fascinating. I can't wait to stay connected with you, Nero, to understand how that continues to develop in the weeks and months ahead. Thank you. I wanted, before I let you go, to ask you, what do you read or listen to to stay current, given the advanced topics you are immersed in on a daily basis, the need for you to look around corners from time to time? Who are the people, thought leaders, authors, podcasters, et cetera, who help you see around those corners that you might recommend? Yeah, look, I'm a big podcast guy. So there are a number of podcasts, I would say. There's How AI, which is on Apple.

25:02There are podcasts from Ethan Mollick I listen to every now and then. I'm a big fan of Ethan. Some great books he's written. The other thing I would say is I'm big about experimentation. So I'm really passionate about vibe coding. So maybe I'll take a few minutes to talk about that. I have probably not written code in probably a decade or not longer. But since Christmas, this past Christmas 25, I started vibe coding. And so I am building things at scale now, Peter, which is really, really innovative. Not for MasterCard necessarily, but I'm staying current. To your point about staying current, it's about getting immersed in the technology.

25:40And so I probably vibe code about an hour, hour and a half every day. building things, everything from home automation systems to playing with the Raspberry Pi. That's my latest passion. I've got an open claw machine at home where I'm running open claw at scale and play with Hermes, which is the open claw competitor. I've got a Mac mini running, so I've got my own local models running locally, play with open wave models and tweaking that. and really, you know, Vibe Coding has really kind of unlocked kind of a new era of enthusiasm for me, right? You know, as a CIO or CTO, you know, you don't get to spend time in technology.

26:19You're spending a lot of time in, you know, technology, but not in the guts of technology, so to speak. Now I'm going back into kind of where I started my career and it's exciting. So I would encourage everyone, you know, to listen to this podcast, really get into it. I mean, it's so easy to start Vibe Coding and it's addictive, right? Once you start building something, you want to build more and more. And I'm building stuff for my wife. I'm building stuff for my parents. I'm experimenting with, you know, the latest aging technologies, which were on the bleeding edge. And it's like, that's really learning.

26:54Podcasts and books are great, but playing with it is even more satisfying. And so, yeah, I listen to podcasts and read some books, but like, this is my new hobby. And I can't encourage folks enough to kind of do it more.

27:08Peter High: Great advice, certainly, Nirov, and a great conversation, more generally speaking. Remarkable the spot you occupy, the perch you have as a result of the role you play in this remarkable organization. And I look forward to staying in touch and continuing to see the great things you and your team bring to life. But thank you very much for the conversation we've had today. Peter, thank you so much. It's been a privilege. Appreciate the time. Thanks for listening to Technovation. If you enjoyed this conversation, subscribe and leave a rating. It helps more leaders find our podcast. Thank you. And I look forward to seeing you next time.