Agentic Commerce and the Future of Payments With Prateek Khamesra

Fred Schonenberg
Hello, everyone, and welcome to the VentureFuel Visionaries podcast. I'm your host, Fred Schonenberg. I'm the founder of VentureFuel, where we help large organizations accelerate growth by partnering with emerging technologies and startups.

Today's guest is someone who loves the complexity behind what most of us take for granted every single day, things like payments. So Prateek Khamesra is a Director of Product at Visa, where he's helping shape the future of payments inside of one of the most global, trusted, and heavily regulated ecosystems in the world. His career path is fascinating, starting off in software engineering, moving through consulting in places like McKinsey, then into financial services, JP Morgan, MasterCard, and now Visa. And along the way, he's worked on everything from digital wallets to buy now, pay later, to large-scale network platforms that operate in milliseconds but impact billions of people. 

So what really stood out to me about Prateek is that he doesn't just talk about innovation in theory, he's lived it within a massive organization or several massive organizations where speed, security, and trust all coexist. And he is a thoughtful product leader, writer, mentor, he was just named to the Electronics Transaction Association's 40 Under 40, which is a huge and well-deserved honor.

We both got to serve as judges for CES this year, and so I'm excited to dive into what caught his eye there and how he stays ahead of emerging technology. And so please join me in welcoming Prateek to the show. Thank you for being here, sir.

Prateek Khamesra
Thanks, Fred. Thanks for the kind intro, looking forward to a conversation.

Fred Schonenberg
As am I, as am I. So I know we both met judging this year at CES, and I think that we exchanged some notes on the fact that it was like this fire hydrant of different types of companies and technologies. And I'm curious, from your standpoint, how do you kind of separate the signal from the noise, right? How do you kind of stop and say, oh, this one is interesting or matters for us?

Prateek Khamesra
That's a good question. I think I usually like whether it's any product at CAC or any innovation within my company. I think that's always one thing that I always look forward to, the first thing is, is it solving the consumer client pain point? And that's the first thing, because you can have a lot of noise, but it's actually helping consumers' pain point. That's the most important problem. 

Is any valid proposition to the problem you are solving? Then as you think about the problem, then you think about how you scale? Because you can have a great product, but if it's not solving and it's not scaling, then it probably would not matter. So it's an appetite to scale.

And the third and most critical part is how easy it is to integrate or implement. You can have the best product possible if it's very difficult to implement, whether it's in any or difficult to use, then it will not be actually having an impact. I think those are the key points that I think I look forward to when I'm judging. And as I'm thinking, I just want to make a quick call, like all the opinions are my own on this podcast. It does not represent my present or past employer.

Fred Schonenberg
Absolutely. One hundred percent. And I am interested in the space in which you have worked most recently, because payments is something as consumers, right, that it feels simple. We tap, we swipe, we click. But there is an incredible orchestration that is happening in the background and it's a space that is moving very quickly. I'm curious. How do you stay on top of all the emerging technologies that are happening outside of your four walls when innovating inside those four walls is already complex and has so many moving pieces?

Prateek Khamesra
That's definitely a good point. I think as you said, it's a very complex ecosystem. But I think since I'm in the power system, I get to see all the challenges that we face. And let's take for example, let's say Fred, you buy a coffee at Starbucks as a consumer. You don't care about what goes. You take your card. You can either tap your wallet, your iPhone or Google wallet or swipe or insert the card within a fraction of a second. Your transaction gets approved. You don't care about the transaction. You move forward with your life by the coffee.

But behind the scenes, there's a lot of things that happen. Just for an example, the moment you swipe your card, there's an acquired behind system that's basically the payment flows from acquired to the network, which is composed of Visa, MasterCard, American Express and US to issuer that basically issues a credit card. And then they do their validation from fraud and risk management and back to the network, back to the merchant.

So there's a lot of four party models involved and everything happened in milliseconds. So one important call out is that within those milliseconds, there's a lot of things that happen. We're doing fraud, making sure that there's less fraud, there's more security. At the end of the day, the aim for any network is to have an approval that we want a transaction to get approved. So behind the scenes, there's scalability. How do we innovate at scale? 

Most importantly, there's stability. What I meant is that the system cannot go down. Imagine your transaction fails because the server is down, it's not a great experience for the consumer. And then most importantly, how do we protect and make it secure so that the consumer trusts the system? Because it's a very trust driven model. If consumers don't trust their credit card, then they won't use it. 

Fred Schonenberg
It's such a fascinating experience. People don't understand all the things that are happening in these very, you know. 

Prateek Khamesra
And just to add to it, I've been in the payment space for more than 10 years. And one thing that I realized when I joined, I think my previous employer, MasterCard, I think when I joined for the first few weeks, I'm like, oh, I was going through something called tokenization, which is nothing but your Apple wallet. I was building that platform. And my first few weeks, oh, I understand this. This seems reasonable, all the innovation that we were doing. 

And as I started going into deeper detail, I'm like, wow, that's so much information. And then there are people who've been in this industry for more than three decades, and they still think that they're still learning something new because there's so much innovation happening every day. At the same time, the system is becoming complex. And imagine like there's trillions of transactions, dollars of amount going every single day by these systems. So imagine the complexity and different scenarios that we have to manage.

Fred Schonenberg
Absolutely. Let me ask you this. So what's the difference between when you started in the industry and now in terms of how these larger organizations are thinking about innovation?

Prateek Khamesra
I think innovation has always been the backbone of any company. I think there's a saying that if you don't become a dinosaur, there's a way because the moment you stop innovating, you become a dinosaur within the ecosystem and you die. So that is not great. And you all have one thing that you have to imagine, like the companies have to innovate because there's always disruption happening within a second. 

So you have to keep going, you don't have to think about the curve, you think about what's next. Do the thing that companies are always thinking about. So from the strategy point, when they've all been the cash cow, any payment ecosystem would have a cash cow, which is their core processing business, like debit, credit card processing. At the same time, you also think about the future. It could be agentic payments right now or stable coin, blockchain, crypto. Like, how do you make sure that you're still relevant in the future? And the next stage of evolution. So then when those things eventually evolve, you're not caught by surprise.

One good example could be when I started in the industry, tokenization was just taking place. What I mean, tokenization is the Apple wallet. So when I think about wallets, whether it's Apple, Google, Samsung, they just started and the penetration was hardly there. But now I would say they have grown almost 10 times within the last few years. And I don't carry cards in a wallet, I just have one credit card and everything else is in my Apple wallet. So that's the experience I see everywhere. 

Every time I walk into a shop, no one is using their cards anymore. They're mostly using their wallets. So that's a change in the behavior in the consumer. So that's a big paradigm shift that has happened in the industry. And it's becoming more digital rather than being card based, which is more plastic driven.

Fred Schonenberg
Let me ask you this question. So along with tokenization, right, it's often easy to look backwards and say, oh, of course, everybody is going to use their Apple wallet. That is going to become a disruptor. But if you go back, I don't remember when it started, five, six, seven years ago.

Prateek Khamesra
2014, 2014, when Apple Wallet.

Fred Schonenberg
So 2014, and you see it's a minuscule amount of transactions. How do you start to determine, hey, this is where we think it's going? Right. And you mentioned some other things, blockchain, stablecoin, there's been fits and starts of some of these technologies where early adopters say this is going to change everything and other people say this is going to go away in five minutes. How do you think about that when you're spotting something that is an interesting, emerging technology, but it's unsure where it's going to go in the future?

Prateek Khamesra
And it's like connecting the dots, right? It's like you really can't connect the dots and you look back. I think one thing that I think this industry has is a partnership model. So the payment ecosystem, as you mentioned, is very complex. It's very difficult to start something on your own. So you can only grow the pie by partnership. So even when Apple Wallet was launched, the Apple Pay, they couldn't launch Apple Pay just by themselves. They need an ecosystem, the partnership model. Obviously they approach all the networks, all the issues that came in launching these new functionalities and which is the Apple Pay.

Now, looking back as I think about the tokenization and Apple Wallet, of course, everyone knew that given the size of the as the more and more people use their iPhone or their Android devices, there's obviously going to be a digital shift. And we already saw that shift as part of e-commerce transactions. And PayPal is a classic example. PayPal Wallet came and disrupted the complete ecosystem back in the Haywire days. Now, Stripe is doing something similar in the ecosystem. So you can definitely see that paradigm shift in 2014 as well when Apple Pay was announced.

And then as it started scaling, it didn't happen overnight, right? It took a long time. But the way I see it is it's the right way towards. You can only say one of those things you can only guess that's going to have some potential, but it only happens when you look back. It's like any other disruption that you can always there's a huge potential that you think that's going to happen. But until those things happen, you can never say with certainty that, hey, and hey, I predicted this thing to happen because like any of the predictions is always in the future. No one knows what's going to happen.

Fred Schonenberg
I remember for a while I was seeing all these different venture capitalists showing me how smart they were and their returns and that they had invested in Uber early. And there were at least 20 that was look, look what I know. I've invested in this one. I'm like, what about the other 98 that didn't work out? And right when you look backwards, you can look very smart, which I think is very interesting. 

So let me ask you this. One of the things the organizations you work in, there's I can't think of any other industry, maybe health care that is more regulated than where you're doing. How do you move fast to seize these opportunities while also knowing that you have this complexity? You have the requirement of reliability and security of billions of people.

Prateek Khamesra
I think obviously people talk about innovation, like why we're not disrupting enough. It's important because when you are a regulated entity, you basically, for example, when you're a financial institution, whether it's a bank or a payment network, you're actually dealing with a lot of sensitive data of consumers, of PII information, right? So it's really sensitive, highly regulated. And you have to think about the consumer. 

Imagine like I always tell people, my mentor used to joke about it, like we cannot be the next startup that will break things because when things break at the company, billions of people get impacted. So one thing that you need to make sure whenever we launch something, we have a thorough testing and it's fine to be late in the game. But it makes matters that you have the right safeguard and the right stable, secure environment for consumers, because we cannot be the next startup where you launch something and it fails and you go back and revert your code. 

You have to make sure whatever you launch is tested well and works. So there's a balancing act there, Fred, I think. What's important is that you need to make sure that you keep the regulators and consumers first. It's basically a consumer first mindset. At the same time, you're also investing in innovation.

So you may not be the first, but it's OK with the second and third in the race in this case, because what matters is that you're making you are in that space, whether it's the next big wave of innovation coming in, because the moment you become first and you try to innovate with this large organization, if anything breaks, it's not only the consumer, it's also regulators and a whole PR disasters out there. Imagine that the network breaks and you're not able to use your Visa or MasterCard or American Express credit card, what would happen? The entire ecosystem or the financial system will crash. 

Fred Schonenberg
I think one thing that's interesting is that a lot of the companies that in the industry you work in have embraced that the startups outside of your company can push the envelope and try these things. And as they hit a certain amount of scale, you can acquire them or partner.

Prateek Khamesra
Exactly. So I think as I mentioned earlier, I think in this ecosystem, partnership is the way to go. So one thing that I think you might have hit the nerve there, I think it's the truth of the financial industry. It's a partnership model. That's the only way to scale for a big company, because any network, whether it's Visa, MasterCard or American Express, cannot scale. 

They cannot move fast enough compared to any startup, which is obviously because startups are much more nimble in space and they don't have regulatory burden on their neck. So they can do whatever they want. But the moment they reach a particular point that's where the partnership model come into place, where like a company like whether it's a payment network, it partners with those companies can get the scale that they want.

At the same time, if the right time comes, it's either an acquisition, which is we've seen in the past happening in the open banking space or in the processing world where and those can those partnership model is a very good way to lead to acquisition, because it's a it's a test of the marriage. Because I'm sure you in the venture world understand it's we've seen a lot of acquisitions failing because you acquire for a purpose and it doesn't work out that way. It's like a bad marriage. So partnership is like testing the water by making sure that it's the right partner. And when the time is right, you basically acquire the company and give them a great exit. At the same time, you get the value that the company is bringing by bringing everything in-house.

Fred Schonenberg
Yeah, I love that. As you look at all the different technologies and trends that are happening within the financial services industry, is there anything that you're excited about? Is there the tokenization Apple wallet of the future that you see emerging?

Prateek Khamesra
I think payment should be the next big thing. So as you know, I think obviously we move from what is AI to everything AI. So imagine like maybe four years back, the chat GPT was launched and changed everything in the world. And today, AI is everywhere. I think with it, I stopped googling and searching things on Gemini or chat GPT. That's the way I didn't imagine the world without Google search.

And here we are now. We right now, the agentic payment is as a space is in the future. I think not even in the future, maybe within one or two years, it's going to be agent shopping on your behalf. Imagine, Fred, right now, what the agent, let's see where we are right now. So you are in, let's say, in chat GPT, you ask, hey, I'm planning to run a marathon. I want to buy a new shoe and give me a recommendation of top five shoes with Hayes McCruddy, like $150, one budget is $150. I need a midsole, I need to be comfortable, blah, blah, blah, blah. 

And you put the intent into the agent, the agent comes back to five options with the links to make payment. You click on it. You have your credentials, your credit card information already present in the chat GPT. You're given the credentials and the agent will go and shop on your behalf. That's where we are right now.

Where I see in the future would be if you give the intent to the agent. Right now, there's still human assistance in place. You are actually doing authentication, which is like you before the payment goes through, you're actually using your biometric, which we call secure card authentication through and making sure that it's actually you making a payment.

What will happen in the future will be that you will give your intent, as I mentioned, on buying the best shoe and the agent will go and look for all the deals, especially around Black Friday. And you're given all the instructions and will basically shop on your behalf. Like you can and they're looking at different deals, you've given the criteria like, hey, this is my budget, this is the deal and you're not even there and automatically you wake up next day and the shoe is on your doorstep.

Fred Schonenberg
Yeah, with the best deal possible, right, that's the next issue.

Prateek Khamesra
Exactly. And I think already there's a great move happening. So think about what exactly that means behind the scenes. What's happening behind the scenes is it's very interesting that we are seeing it's nothing but data talking to data. Just to explain you an example, like what merchants started doing is that they started sharing this cute information, which is nothing but their specifications, their inventory to the agents. 

So our agent automatically knows what the specifications of the shoes are and how much inventory the Amazons and Nikes of the world have. So now basically it has all the right information already integrated into the API. And now all it's doing is basically acting on behalf of the consumer. Once a consumer gives the intent and the consent. Those are two important things, the intent and the consent in this ecosystem. 

Fred Schonenberg
One of the things that I think is interesting to take your example a little farther is we were talking about GEO or AIEO, all the different versions of what search engine optimization has become. The idea used to be on Google. And I shared this with a very large advertiser and they said, it's only a small percentage. We're going to be Googling forever. That was 18 months ago and now they are switching their song. And so they're very focused on this space. And what we shared in the next stage is essentially what you're saying is this agentic commerce.

Prateek Khamesra
And yes, that's the next big thing. And it's already started. And I think we started seeing volume. Obviously it's not big because, again, the adoption is slow. Like every technology, it's early in its stage. So there are still early adopters, which are mostly big players, which can afford to invest in the technology.

But I do see it happening the more it's going to be the more consumers rely on using the agents for every day. And I think at one point, I believe the agent will know more about your choice than your partner. What color you like, what is your shoe size and what size clothes you fit? I think because you're going to feed all the information to the agent and it's going to remember it. Maybe six months ago you mentioned you like the color blue. So it already has the information captured. So now it knows exactly what you order based on your preferences.

Fred Schonenberg
What's interesting is one of the things that these companies are worried about is this idea of data to data. We've talked about bot to bot and the idea that let's say I want this running shoe. And truthfully, maybe I care that it's blue, but I care less about that than the price and the agent knows. And essentially, the agent's job is to go negotiate down the other agents or to find the deals. And so one of the concerns is that it becomes a race to the bottom, right? Where profit margins shrink, brand has no value. 

But on the consumer side, I love it because my hundred and fifty dollar shoes, I got for one hundred and twenty. And so it's interesting. I never thought about this conversation. I never thought about the payment layer.

Prateek Khamesra
And think about it, Fred, like right now, when you shop, right, it's obviously a human cycle. The agent does not have human psychology in place. Like you go to Nike, even before looking at the shoe, you're basically looking at the advertisement, whether you see athletes and your favorite athlete running or it's the visuals. 

The agent doesn't have anything. All they care about is what intent the consumer has given it. You just focus on that. And what is the intent for you? It could be the one fifty dollar budget. It could be blue. What is your preference? Yeah, that's all it cares about. And maybe your budget is flexible. You can say up to two hundred dollars. But for me, comfort matters more. So obviously it's going to take priority over the budget. So whatever instructions you will give, the agent is still all that human behavior which comes into picture from the marketing point of view.

But you think about it like, hey, if I put a great logo, great image, those things will come up. They are still there. The brand value still remains because the brand has some quality. But it might diminish if the agent is doing shopping on behalf because it's looking for comfort, looking at what the instructions the consumer has given it. 

Fred Schonenberg
I had a very interesting conversation about this idea that certain brands have studied human psychology for hundreds of years and they know if they put the ad on the end of the aisle and it's red, it'll drive this behavior. And they asked me, how do we do that with AI? If the bot is shopping, we want to basically persuade the bot to make a psychological decision. And I said, I don't think you're thinking about the right way. It's going to go after the intent of whoever is representing as an agent and your red color isn't going to do anything unless that's important to the person specifically that they prescribed it to be important.

Prateek Khamesra
That's correct. And I think one thing I'll say right now, as humans, we're really good at marketing and we're getting better at it every day. That's the fact of life. I think that's time tested. I think that's one thing we and that's where I think one thing I think slowly I feel that that's going to come into. Obviously, they will be sponsored later. I think that's where because of the amount of running these agents have, they cannot sustain without the ad revenue.

So slowly, what you start seeing is that the intent would matter. But I think when you search first, how Google search will get the first two is always paid. Similarly, like in the agent e-commerce world, I think first to search will obviously through paid advertisement. And then the next three would be the actual search. So they would be. I see advertisements coming to picture because the revenue model otherwise is not sustainable for any of these payments, like any of these agentic companies because they're billing billions of dollars a month on just running the GPUs.

Fred Schonenberg
Yeah, at some point they're going to have to make money, right? They can't just keep raising, can't just keep raising.

Prateek Khamesra
It's right now, it's venture money. Someday it's going to, well, there is a saying that the well is going to run dry someday. So you want to make sure that you have a plan for how to make revenue. We started seeing ads in OpenAI and Google is launching ads next year. So they've already announced it.

Fred Schonenberg
Yeah, it's certainly coming. So let me ask you this, I'm going to throw just a couple of buzzwords that I know from your industry, and I'd love you to give me a one sentence thought on what it is or the future of it. So blockchain, are you a proponent? Do you believe in it? Where do you think it's going? 

Prateek Khamesra
It's still very volatile. I think blockchain, so I'll give to the crypto and blockchain are different. Blockchain is the mechanism that drives crypto, right? So as everyone, crypto, people have been saying the future for the last five years, have yet to see the future. But given the nature, unless the regulatory framework around the whole crypto, I don't think so it's scaling.

And again, that defeats the purpose, the moment you bring regulation that defeats the purpose of crypto and blockchain. But I do think the stablecoin has some significance. That's a bifurcation, stablecoin is nothing but more of a regulated functionalities. It's again, P2P, it can be destructive, but we haven't seen much transaction happening there. But that's the next big thing in terms of blockchain, that could be the next big wave. If you think about blockchain.

Fred Schonenberg
Yeah, interesting.

Prateek Khamesra
But as a concept, blockchain is amazing. I still think there's a huge value from the concept point of view, but it's one of those things that the application is always a little clunky.

Fred Schonenberg
I love it. Yeah, I mean, you just nailed my three that I was going to ask you, I was going to say blockchain and crypto. So I think we covered them there. Is there anything beyond agentic that you think is really interesting for the future of currency, payments, anything that you think is bubbling up?

Prateek Khamesra
I think obviously agentic payment is what everyone feels. But I do see there will be a lot of operational efficiencies that AI will bring into the ecosystem. What I mean by operational efficiency is as I alluded earlier in my conversation, the networks are processing trillions of transactions every day.

So with that, I think there's a huge opportunity for AI to help optimize whether it's less fraud, making sure that we are one step ahead from the fraudster, and making sure that the approval of the transaction happens efficiently. And the most important, making sure that the system is secure and stable. But that's the only way this ecosystem will work.

And that will make sure that the trust of the consumer remains intact. So I do see a lot of operational efficiencies that this AI will bring in. It's already as I'm already seeing a lot of live examples like my company where we're investing heavily in those systems. But I think moving forward, I see a lot of big changes on that part. And no one talks about the operations side because that's not the sexy part of the work. But that basically keeps the machine running.

Fred Schonenberg
Of course, it's very important. Well, Prateek, this has been amazing. Thank you for taking the time.

Prateek Khamesra
Thank you. Thank you, Fred.