Rezolve CEO Sees Growing Competition Over Control of AI Shopping

Narrator 0:06

This is Monday Conversation, a PYMNTS podcast. Karen Webster sits down with the visionaries behind the trends for the stories shaping what's next in payments and commerce. In this episode, Payments CEO Karen Webster sits down with Rezolve Ai Founder and CEO Dan Wagner to talk about how Agentic Commerce meets guardrails around inventory, identity, and payments.

Karen Webster 0:38

Dan Wagner is the founder and CEO of Rezolve, a company focused on bringing AI-driven commerce directly into the consumer experience. Long known as a bold and sometimes controversial voice in technology, Dan has built a career at the intersection of data, digital transformation, and disruptive business models. From his early work in data analytics back in 1984 to his current focus on embedding AI into the fabric of commerce, he's consistently pushed the idea that technology shouldn't just inform decisions, it should also execute them. So today we're going to explore what that means in the age of agentic AI. If software can act, transact, and make decisions on behalf of consumers and businesses, how does that really reshape commerce, control, and trust? Hey Dan, it's great to see you. I'm looking forward to getting caught up on all the great things you're doing with the Resolve and in particular your points of view on the future of commerce when agents are at the controls. So thanks for taking the time.

Dan Wagner 1:45

Thank you very much for having me. It's a pleasure.

Karen Webster 1:48

So let's start with a little context. You've spent a long time really trying to reshape how people interact with commerce and data from your very, very early days in 1984. That was that was sort of groundbreaking with MAID MAID. I mean, to start us off, in your view, what's changed the most over that period of time? What keeps breaking? And now with Agentic, whether you think commerce and innovations related to commerce is really different.

Dan Wagner 2:31

Well, so when I started, uh people didn't have computers. So the IBM PC was being introduced on the year that I started my first information business or my online business. And so what's happened in that period, you know, from the the the early days of technology making things easier, uh from you know, having in my case, having to go to a library. Uh and in 1984, I started digitizing newspapers so you could interrogate a newspaper and get an article instead of having to go to the library to get it physically. Um, that seems like an age ago, and it was, um, to then being able to search anything on the internet 10, 15 years later, and then being able to buy anything without leaving your home. Um, these generational changes in in the way in which we interact with the world around us uh has happened in a in a fairly contracted period of time. But what's happening now is in a whole new scale of extraordinary potential. I've never seen anything quite so exhilarating uh as what's happening with generative AI. And of course, people like me were very much involved in its nascent development, you know, from concept-based retrieval technologies in the 90s, uh, where, you know, to predictive typing. This is all part of the grounding of uh what's known as Gen AI, where the perception is you're having a conversation with a computer when what you're actually doing is talking to a probabilistic mathematical algorithm that is guessing the next word, the next word, the next word to make a sentence and the next sentence to create the impression of a conversation. But for all intents and purposes, it is a conversation because that's how we communicate. I think that the the implications of that are much greater than simply the convenience of conversing with computers in the way that we converse with human beings. Uh and I have a view around how this is going to affect our how this is going to uh uh uh expose itself to us in changing the way we interact with the internet as we know it or digital channels. Uh and I believe it's going to be all agentic, and it's gonna be agentic, and what I mean by that is a consumer to an agent or an agent to an agent. So to give you the example, I go www maces.com and I have a conversation with an agent, a technology, a piece of technology that represents Maces as the best salesperson for Macy's that you could possibly meet if you were to walk into Macy's tomorrow. And they would walk you around the store, get whatever you need because they know everything about every product. That's the kind of uh thing that we're delivering to customers.

Karen Webster 5:41

But but but but but but but Dan, that assumes that you're going to www.maces.com, which a lot of people don't start their journey at the merchant, they start their journey at the prompt with one of the AI models, which is so that's the second, so that's the second uh that's the second one, Carol, which is agent to agent.

Dan Wagner 6:02

So the second scenario is that Dan Wagner doesn't go to Macy's.com instead of going to Macy's and saying, hey, I'm looking to buy a gift for my wife, you know, I don't want to spend more than a couple hundred bucks and I don't know what to get, and being presented with lots of different products from different departments instead of me going department to department, the way we do it at the moment is not really the way you shop online, you know, the way you shop in physical stores is different. Uh, we're gonna replicate through generative AI, and we are doing that, a much richer experience. But in answer to your point, you're absolutely right. The other way is that I go to ChatGPT or I go to Gemini or I go to Siri or I go to Copilot or something, and I say, Hey, I'm looking for a gift for my wife. I don't want to spend more than 200 bucks, you know, I'm not quite sure what to get. What do you recommend? Now, what happens? It's not happening yet exactly, it's started starting down, but what will happen is OpenAI or one of these agents, one of these uh interfaces will fire off hundreds of agents to hundreds of retail stores in a blink of an eye and interrogate all of them and then assimilate all the results and then come to me and say, hey, there's lots of things that you can buy, your wife that is under 200 bucks, from these things to these things. Or how can I help you navigate that down? Uh uh and eventually I will end up buying something from a retailer. Okay, so there's two things that that that implication that implies. The first is that you and I, if I said I've got to go and buy a pair of trainers, I'd go WWFootLocker and WW Nike, I go to one or two stores, maybe via Google or whatever, and then I'd end up selecting a pair of shoes and buying them, right? But the future is gonna be that I'm gonna go to ChatGBT or Gemini and it's gonna interrogate 500 stores. Every time I ask for something, instead of it being two or three, it's gonna be 500 in real time. The implications of that are that every e-commerce platform on the planet needs to scale up to manage that load. That load and the internet itself needs to scale up to manage that load. Is the infrastructure there? I don't think so. So we're investing in an infrastructure in the blockchain. I I think that there is gonna be dramatic shifts here that a lot of people don't fully appreciate. Even people in my industry who are talking, talk, agentic, blah, blah, they don't even have a clue of what the implications of this will be.

Karen Webster 8:27

So I agree that that is the future, and we've seen some of the early evidence that you know the version one pilots of of what you described, the true eugenic experience, haven't necessarily delivered the outcomes that everyone's hopes and expectations thought they would deliver. I wrote a piece on this a couple of weeks ago. I called it my toaster problem, where I did exactly what you what you suggested, but I I knew the toaster I wanted to buy, but I was sort of testing the models to see whether they would actually return the brand of toaster that I wanted. I wanted to see how extensive the agent search would get, and then what would happen once I was given a selection of things to buy. And the toaster problem, you know, failed on two, on two things. One, it didn't return the brand that I wanted to buy. And two, I couldn't buy it. I couldn't buy a toaster from the agent.

Dan Wagner 9:25

Yeah.

Karen Webster 9:26

And so now, you know, there's the number one thing that people are doing with these models is search and discovery. Like your example, I'm I need to buy a birthday present, an anniversary present, a baby present. Tell me what you know the hot items are. But it's a it's a smart search function today. It's not a commerce transaction because you can't complete it. So how far away are we from solving my toaster problem and your query to buy your wife's birthday present from actually happening in your view, based on what you know now, and based on where you think we are from an infrastructure protocol, merchant readiness, et cetera, perspective?

Dan Wagner 10:11

I think we're not we're not particularly close. I think that there is a fundamental disconnect between the hallucinations that these models are inherently uh pushing out and the concern from retailers to open the doors to their warehouses to these agents to run amok and represent their products to their customers, to these end customers without any control or constraint. What we're doing, and I think we're the only ones, I honestly think we're the only ones in the market doing this, right? Is that we are we have built agent infrastructure for the merchant to talk to agents, to control the narrative with the agents that are coming into their site. Uh, so that when the agent comes in, instead of the agent coming in from Chat GPT, taking everything off the shelves and misrepresenting it to me, the customer, the agent comes into the retail site and says, I'm looking for a pair of trainers size 10 in white. And we and the and the agent on our side says, Whoa, whoa, who are you talking to? Why do you want the what trainer do you want? What's the what is the query that was asked of you to come to the conclusion that it's a trainer size 10? You know, because the problem is that these other engines make mistakes, terrible mistakes. We've seen it recently with Walmart kicking out OpenAI and many, many other examples you would all know. And the reason for that is that the old the architecture of these technologies are not fit for purpose for this particular role. And we spent 10 years perfecting it so that it is. And so we want to be the gatekeeper. The best person to sell trainers to a consumer is the brand owner, Nike, or Footlocker, who've got great knowledge about the products, but not ChatGPT, who's a generalist. And especially when you've got this fundamental problem with the technology making mistakes, hallucinating.

Karen Webster 12:13

I I agree with that. And and you know, we we've seen somewhat of this movie in real life today with Google Shopping, right? So you go into a search bar, you do you know, Google, I want to buy you know, pair of black kitten heel suede, size 35, give me the selection, and then you click through to a site, and you know, if you're lucky, you actually see that there's inventory in your size. Often you don't, because it's not real-time inventory management. You just go to the product page and then you go from there. But but what does the agentic future, in your view, as you see it and as you're solving for it, do that creates the brand affinity, sort of the merchant, you know, the merchant um relationship with the consumer, but also gives the consumer the optionality that they obviously want when they're sitting down at the prompt.

Dan Wagner 13:11

So that's a very good question. Right now, Perplexity, OpenAI, Gemini, and others are pushing their standards to retail and say, open your doors and accept this uh uh standard so that we can come in and really, yeah, we can pull all your products off the shelves with live inventory data and everything else. And a lot of readers are going, no, not doing that, because you don't know my products, and you're I don't you're not gonna disaggregate me from my customers. But that's actually the wrong answer. Because the right answer is the consumer doesn't want to deal, like you said, the consumer wants to go to Chat GPT and ask the question. And the consumer really wants to end up completing the order through chat, they don't want to end up going to a site, they want to be able to complete that order there, right? But in order for that to happen, where it meets the objectives that the merchant has of knowing its customers, serving its customers, putting the right products in front of its customers, which it knows better than anybody else, is to have an energetic interface on their side that controls the narrative. That's what we do, right? And that's but we're the only ones right now, and it's too early stage. It needs to be common practice. And you know, obviously, we'd like to be the only supplier, but it's not going to be that way, right? That each merchant has the infrastructure to support the agents that are coming in, control the narrative and allow that transaction to take place, knowing who the end customer is, that that end customer is a regular customer, a premium customer, has loyalty, doesn't have loyalty, you know, likes this kind of room, not that kind of room, whatever it is, all the things that the merchant knows about their customers or knows about that type of customer that may be appropriate for this other type of customer that ChatGBT and Gemini do not know. They cannot, they might think they can stand in front of everything, but they can't represent the merchant properly. So the merchant needs to be. I think that's partly why you're seeing such amazing momentum in our business, because we talk sense to the merchants. We say, look, we're going to solve this for you. You need to be visible to those guys, but you need to control the narrative. And that's sort of that is how the future of agenda commerce will play out. A symbiotic relationship between the merchant and the front end of Chat GPT or Gemini or whoever it is.

Karen Webster 15:27

Right. I mean, the merchant wants to remain the merchant of record. They don't want to be anonymized as part of these platforms and the aggregated platforms.

Dan Wagner 15:37

I I don't think that that swap is being promoted by the other players. I don't think they're trying to be the merchant of record. And if they are, that's never going to work, right? But I think that they want to have access to insert the order into the retailer's site on behalf of their customers, you know, just because they did a deal with Stripe, I think ChatGPT, and they did, you know, we'll pay, we'll pay you, we'll sort it out, we'll push the order into your systems, you know, you then deliver it to the customer. I think the merchants are going, whoa, hold on a sec. You know, we know our customers, we don't want to just have these orders come to us that ultimately might be having to be refunded or replaced because it was sold incorrectly. We don't want that overhead. And we don't want to lose connection with our end customer. We want to maintain that experience.

Karen Webster 16:23

I think it depends on the business model. You know, are these models a marketplace or are they simply another commerce channel? And I think that's we don't know the answer really at this point, because if you think about the potential for what you described, Agentic is a completely different way of enabling commerce, um, whether it's completely fulfilled in the in the virtual digital world or whether it's started in the virtual world, digital world, and fulfilled in the physical world. But but it but the way we're talking about it today, it's almost as if it's another digital channel. And I think that's short, I think that's shortchanging its potential.

Dan Wagner 17:09

You're absolutely right, Karen. You are a thousand percent right. And it is absolutely equivalent to the file. I keep saying it. E-commerce as we know it is dead. It's over. That interface is finished. That experience that we have of searching a search box and navigating with a left-hand chat set of categories and getting 40 items and 40s and going through a check. That's finished, that's never going to be the case again. That's like mail order, telephone mail order, mail-in, mail order, right? Where you get a catalog and you look through it and you pull out a coupon and you send it back in the mail, which you probably won't remember, you're too young. And it was around and very prevalent in the 50s and 60s. And now we have a situation where e-commerce is over. Nobody's going to shop that way anymore. The new way needs a new system. The new system is agentic, the new agentic system sits on the blockchain, the new payment systems is stable coins. The infrastructure has to accommodate volume of activity that is hundreds and hundreds of times greater than what we're seeing today. Because agents don't go one by one, they go everywhere instantly. And so volume is going to go through the roof, interaction is going to probably remain the same, but the activity behind the scenes are going to be massively more uh uh uh uh um intense. And the infrastructure that we have today doesn't support that.

Karen Webster 18:37

So, Dan, what's on the critical path as you envision agentic as this next evolution of commerce? What needs to go right? And what could stand in the way of that going right, in your view?

Dan Wagner 18:55

Well, I I don't think I think that there needs to be a new payment infrastructure in place that replaces the existing one. And of course, we've introduced Resolve Pay, and we believe that is it. Okay, so I'm doubling down big time on that. I've got an infrastructure sitting behind it in the blockchain, stablecoin payments, and so on. And and although it's not uh in the public domain yet, a very innovative and very cool way to onboard millions and millions of consumers in a very elegant way, which is not there today by anybody else, right?

Karen Webster 19:25

So so so why so why do you think there needs to be new payment credentials when consumers have payment credentials that are tokenized, and these tokens can carry intelligence around identity and preferences?

Dan Wagner 19:40

You can't no, that's possible. You can do that. It does, but it's it's a it it it's it's old tech. It's it's sitting on old rails. And uh the the end payment is the volume of end payments isn't going to go up massively unless you're using micropayments, which of course uh um a gen tick payment infrastructure can support uh and stable coins can support, but traditional currencies, uh fiat currencies can't, because you can't pay with a fraction of a penny somewhere. So the old roles aren't designed for the new era. Will they be around for a while? Yes, they'll be around for 20 years. I don't think they're gonna go as quickly as e-commerce will go. Um, but I think that the uh I think that one of those, one of the tenants is payment is a payment infrastructure designed for the next phase, and the other is a commerce infrastructure that's designed for the next phase. And our commerce infrastructure is not designed today for the next phase. So we we are trying to change that. And remember, I was one of the pioneers in um in online enterprise e-commerce with with Bender in 1998, which was a cloud-based commerce stack before we were thinking of cloud-based anything. Yeah, yeah. Uh that was way that was way before people were thinking about cloud. It was, in fact, uh Salesforce started in 1999, we started in 1998. So we're all kind of right at the lead of that SaaS revolution. And here we are now at the lead of the agentic revolution, and it is, I mean, infinitely bigger an opportunity, infinitely bigger um uh a transition. And it's it's super exciting because the retailers that get it right, just as they did in the dot-com era, the retailers that get it right will run away with this. Because if they can this, if they can represent their products properly through the channels that people are going to go to naturally, Siri and Google, Gemini and Chat GBT and what have you, uh, and Alexa or whatever, uh, and if they can represent their products there properly, non hallucinatory and you know, not not conflate, and they can manage the agent volume that's going to come to their sites, interrogating their sites and answer them in an intelligent way, manage them in an intelligent way, will win. They'll win.

Karen Webster 22:00

Who's leaning in from the merchant perspective? You don't have to give names, but are the enterprise middle-tier merchants? Who really says, yes, I get this, I understand it. I want to be part of the first wave.

Dan Wagner 22:17

Well, I think everybody who's coming to us, and we've seen this, yeah, I'm sure you know, under unbelievable momentum. And I built businesses, as you know, through my career. I've never seen anything like this. This is, you know, extraordinary. I mean, it's gratifying, but extraordinary. I think that a lot of these retailers want to be with the company that's holding the the beacon, showing the way, right? We're showing the way. We're saying this is how it's going to be, but I don't think anyone's talking like us. There isn't anyone out there who is setting out a path. There's a lot of kind of, oh, the stablecoin, you know, we're gonna there's oh agentic. I don't think anyone's actually, they're just throwing these names around. I don't think there's a real clarity around what I've been saying to you. I think I'm telling you exactly how it's gonna be, and there's nobody out there who's saying it anything else, right? They're just talking agentic, who's really saying we're doing this or that to meet the market demand. This is how it's gonna go, because I don't think anyone's being bold enough to stand by their convictions in this. I think there's a lot of noise out there, and that's fine because it typically happens in these situations that you get this kind of dynamic. But um we do believe that we know exactly what's happening, and we do believe what needs to be done, and we're doubling down on making sure that we take advantage of that knowledge. Uh, and I think that the market, even if we're wrong, I don't think we are, of course, but even if we're wrong, the merchant base that we're signing up want to believe us, want to be with us because they so I think the answer to your question is slightly caveated. It's I don't think that the retailers who are who are with us today are all into this. I think they're just all into our vision. I think they're all into the fact that they need somebody to give them a guide to this because they don't know which way to turn. And we're the only ones who have a very clear, credible path to navigate through this.

Karen Webster 24:14

Well, listen, Dan, I hope we have the chance to uh to chat again. Congratulations on all the great work that you've done at Resolve and you know look forward to staying in touch. It is a very dynamic, active, and very interesting space.

Dan Wagner 24:27

Karen, thank you very much for inviting me to be on this call and very much to be part of this conversation. And I look forward to speaking with you again.

Karen Webster 24:33

Thank you, me too. Bye-bye now.

Dan Wagner 24:35

Okay, bye.

Narrator 24:40

That's it for this episode of the PYMNTS Podcast: The Thinking Behind the Doing. Conversations with the leaders transforming payments, commerce, and the digital economy. Be sure to follow us on Spotify and Apple Podcasts. You can also catch every episode at PYMNTS.com/ podcasts. Thanks for listening.