What happens when AI agents become customers?

0:00Here's the weird thing. We have customers knocking on our door like, we need this now. We're like, what? Okay, really? What does that look like? From day one, when we looked at Solvapay and thinking about how do we do agent-to-agent transactions, like the fully autonomous, or even thinking about it, human-in-the-loop, kind of hybrid, and then the fully autonomous thing is a little bit more sci-fi. That's at least a couple of months out. You have agents now that can analyze, it can write code, it can think for you, and it can act, but it can't really transact yet. Two things to that. We bring identity, often kind of security, who runs that agent, but it will also give them the ability to transact, give them means of spending.

0:37We cover both sides of that as this agentic economy is taking shape. I think Vigo saw in my face and my reactions that I thought, wait a minute, that's actually pretty huge because there are no payment rails for the agentic world. The more you think about use cases, the more excited I got.

0:58Andreas Munk Holm:It sounds to me like we're looking into a future where your next customer isn't going to be a human. It's going to be an agent. Welcome to the European VC podcast. For two years, we've debated what agents can do. The unglamorous answer is that most of what they can't do yet is blocked by payments. Cards and Swift were built for humans typing in their CVV. Stablecoins sound clean until you try to plug them back into the bank account that pays your salaries. Stripe is moving, Coinbase, Visa, and MasterCard have all announced. And out of Stockholm, Viggo Steinsdorf and his co-founders quietly built what they pitched as a neutral cross-protocol layer between all of them, buy-side and sell-side, sitting on the rails that already exist.

1:35Andreas Munk Holm:In this episode, you're getting two segments. First, Viggo and me, the founder caught, what Solvapay actually is, who's buying it, and the thing that caught Viggo by surprise post-launch. Customers banging on the door, asking for fully autonomous agent-to-agent transactions right now. He thought that was 12 months out, but it clearly isn't. Then we bring in the lead investors, Michael Bilas, general partner at Redstone and the fintech anchor there, and Samuel A. Siren, Redstone's CEO and founding partner. Why Redstone went in this early on infrastructure, most LPs would still call too speculative to underwrite.

2:08Andreas Munk Holm:What they saw the closest for them and whether they think agentic payments end up winner-take-most or a coexistence layer. If you run a SaaS or an API or a data product, this one is for you. Your next customer probably isn't human. Let's get into it.

2:27Andreas Munk Holm:This show is not investment advice, and the hosts of this episode may be invested in the funds and companies featured. Tell me exactly what are you building? Great. We're building the payment layer for the agentic economy, taking care of both the buy and the sell side. What does that mean? Well, it means that if you are a merchant today, you have a SaaS digital service, I go out and say, you know what? There's a new customer that is looking for you. And it's not in need of a new pair of shoes. It needs your data set. It needs your API. It needs your specific knowledge and analytics tools that you can provide.

3:05We enable you to sell that to agents in agentic systems. And then on the other side, you have the buy side. You have agents now that can analyze, it can write code, it can think for you and it can act, but it can't really transact yet. So we bring two things to that. We bring identity, often kind of security, who runs that agent, but it will also give them the ability to transact, give them means of spending. So So we cover both sides of that as this agentic economy is taking shape. We have a lot of players in it and we can dive into specifics of it. But it's really like enabling both the buy and sell side here.

3:49Andreas Munk Holm:And who are the buyers and the sellers? If you should describe the ICP here, is this Amazon for agents? In other words, is anyone who has something to sell and is anyone who has something to buy? Great question. We're seeing some patterns emerging. And of course, this is where we are right now. So builders, we have a new wave of builders. Anyone can be a builder. You can build a product now and the cost is going to near zero. You can go from prototype to production, and now you're going to go to paid within seconds as well. So that means for us, it's also building a payment solution that enables builders.

4:32You don't have to do a lot of configuration. it's add a skill, add Solvapay as a skill to your agent and tell it to integrate into whatever app you're building. So that's a part of it. At the same time, we also want to enable that builder and the next ICP, the SaaS owner, for what is coming. You have agents, they need your data set, they need to be able to transact on it and pay for it. So there are multiple ways we can enable that. We've really focused on like a cross-platform, cross-protocol level. So yes, you have your old sales funnel. You're selling a subscription over here. Let's enable that parallel thing.

5:12The agentic new revenue stream where an agent on behalf of a human or autonomously can find your data set, try it. And if it likes it, you can actually buy it, buy more credits to use it, etc. So that's really like enabling the agentic sell side for the sauce. And the third thing that this actually caught me from like by surprise, because I thought this was further out, to be honest. From day one, when we looked at Solvapay and thinking about how do we do agent to agent transactions like the fully autonomous. So we were thinking about it, human in the loop, kind of hybrid. And then the fully autonomous thing is a little bit more sci-fi.

5:50That's at least a couple of months out. And here's the weird thing. Like we have customers knocking on our door and like, we need this now. We're like, what? Okay, really? What does that look like? There's a common thing there. And maybe someone will recognize themselves in this. It's like platforms. You have maybe a creator economy. So one of our early talks is with a client that has creators building purpose-built data sets, dashboards, analytics tools. and they surface them through their platform. So this customer of ours, and they have a rev share model and they want agents to be able to spend autonomously to do analytics and use these different creator built tools.

6:37And they're like, can you solve this? And we're like, yeah, that is actually a prime use case for an agent to agent. So that really caught us a little bit by surprise that they are actually thinking about this already.

6:49Andreas Munk Holm:Can you tell me, is this going to run on blockchain or does it plug into the existing financial rails? For us, it's crucial that it fits and it plugs right into the existing financial rails. We're not anti-crypto. Stablecoins is getting adoption. You know, it's a little bit tempting to only go that route. We saw some other companies that went that route because you can just sidetrack a lot of things. But once you have to bring it back into reality, you hit the wall. You're still seeing that the world isn't there. The world still runs on SWIFT and like the old traditional payment rails. And if you go fully crypto, it still doesn't fit in.

7:30So for us, it's fundamental what we build works in the existing financial world. And it's important for businesses, I would say, because if you introduce risk and you have a service that is like not connected to the banking that you're using, you're introducing risk there. And that's not us. We have to be enablers, not adding risk to that system.

7:57Andreas Munk Holm:Tell me, I want to ask you a question here on the wider, agentic economy. I'd love to hear your take on where are we? How fast is it moving? As you said, you thought that there were use cases that would only be needed in at least a couple of months. We used to talk about years, but at least in a couple of months. But you've seen use cases pop up and that really just needs to be out there deployed right now. Can you talk a bit about adoption across the startup market? And I know that startups are a great ICP for you because they're fast adopters. They need this in a faster way than the incumbents.

8:31Andreas Munk Holm:But I'd love to hear what are your conversations with incumbents who are moving fast, who are saying, SolvePay, I got to know what you're doing and I got to figure out how we can get that into our banking system? Quite a bit to unpack there. There are a few things that I can highlight. I call it, lastly, like the battle of protocols. So we have a lot of actors in this space. We have the LLM providers that have clients, and they're now building out their own marketplaces. We can think about them as their own bubble. So you have a lot of agents in here, and they have services being surfaced, but they can only operate in this bubble.

9:06So if you have like OpenAI and Hazard Marketplace, you can't really like go outside of that and buy things in the Anthropic Marketplace. And how does that relate to like the battle of the protocols? But what we're seeing is both the financial layer and the technology providers are like building out protocols. Some are more gatekept, like purpose built for this bubble. Some are more universal and some are now being adopted and moved out. And like, so MCP is an example. It differs than other like payment protocols specifically. But if we think about it from the perspective of how can we make something that is universally connectable that can span this ecosystem, that is something that is really important for us.

9:52So if you have a business, you want to be able to sell and operate in all these bubbles, right? So who wins the protocol battle? We don't know. I think the universal protocols and the incumbents that embrace that, thinking about the USB standard for this, is what's going to drive mass adoption. And that's how we see it as well at Solve-O-Pay. We want your business, connect with us, and we'll make sure that you connect across all these different protocols so you can sell and connect to all these different bubbles.

10:28Andreas Munk Holm:Nico, before we close, I just want to ask you a bit about your roadmap. Tell me where you're headed, or what does the next 12 months look like for you? So it's a speed run of getting regulatory approvals. So, I mean, we truly want to have some licenses that enables us to do more for our customers and actually go down on the transaction layer. Right now, we have great payment rails underneath, and we can do things like that. So that's a big part of it. So adopting and being that multi-pronged approach where we connect you as a business to sell inside of all these bubbles is a top priority for us.

11:05We have some cross protocols, but now we want to venture in and do deep integrations inside of all these bubbles. So for the next 12 months, that means that a business connecting and using our SDK or even a no-code solution that we have for MCP servers, that means that you can build per tool call, etc. will be able to sell and operate in these new marketplaces that are taking shape.

11:32Andreas Munk Holm:It sounds to me like we're looking into a future where your next customer isn't going to be a human. It's going to be an agent. And for that reason, I want to close this really quick announcement episode by saying to everyone out there that if you have a software as a service product, if you have an API or a data set, it's already being scraped by AI. or if not, it's being unused. And that is quite a terrible state to be in. So go to check out Solvapay and figure out how you can monetize your product in the authentic world. Solvapay.com is your place to go to. That was a bit of a commercial. I wanted to, at least because we have so many people tuning in that are exactly your type of use case.

12:12Andreas Munk Holm:So I think that they should definitely be checking out Solvapay. Samalee, could you just kind of provide everyone with the context of where we're at when it comes. You've invested for decades, I was about to say, in fintech. So you've really seen the shifts that we've gone through and been there at the forefront. I'd love to understand from you, we're all talking about the agentic economy. Is that just hype? Is it a real word that matters? How do you see it? Well, it feels like a hype, but it's quite a bit more than that. So it is a real thing and it's for sure happening. And it's always, there's a bit of the overreaction, And like every time, you know, as the internet started, new economy, then we had the web 2.0, we had the blockchain happening.

12:56All these things were there and quite a lot of that stayed and was as amazing as we always thought. But of course, there's quite a lot of bubble up top of that. And this is the same thing here. I find it very exciting. Mikael knows it well. A few days before ChatShift was released, I said, everybody, guys, now there's a big thing coming and happening. And now we've been experiencing this for two and a half years. And that's why we invest in that, because we believe that quite a bit of this AI agentic stuff is definitely happening. At the same time, it's very challenging to figure out what is really long lasting, what is really a game changer, what is more like, let's say, a discussion out there and horror scenarios and everything.

13:40I see it very positive and quite a lot of interesting stuff happening.

13:44Andreas Munk Holm:Question from the sidelines here. Can I say, Samuel, that you guys also invested in Grok and maybe just ask you to say a little bit about that? Yes, we did invest. And now it's, of course, exciting times because Kroc is the ex-AI and now we are through the merger shareholder of SpaceX. So crazy, right, in general. And yes, we did invest underlying LLM because we said there are going to be only a few big winners who really brings this enabling technology. And that's why as the opportunity was there to invest as a seed investor into that, we took it. And it's been extremely exciting, extremely fast moving.

14:23There have been ups and downs on that topic on multiple reasons. But now it's turning to become one of maybe the best investments we've ever made. So let's see how the IPO goes.

14:35Andreas Munk Holm:I just wanted it in there because I don't think most people know that you guys were actually in that round. I think it's an exciting and important thing, both for Redstone, but also, of course, the European venture community. because I don't think that people entirely understand how well networked and what access many European firms actually have. Now, Vico, I want to ask you, because I just asked Samali just before to describe kind of how he's seeing the market. You're, of course, directly at the transaction side, and I'd love to ask you, are you seeing what Samali just described showing up in the data, showing up how your customers are using Solvapay and so on?

15:15great question um is it hype or is it is it real um as a founder i find myself waking up every morning i'm like oh i'm late and then uh right in the next call i'm like oh we're early and that to me is a signal that we are actually right at that infliction point where this shift is happening and if there's something that is hard it's changing human behavior. But this is something new. This is something that is acting on behalf of. There is a component that just has sped things up. That is, like Samuel said, headed towards something that is catastrophic. It could be. But there's so much good happening in it as well.

15:58There's so much things that use up human potential and time that just can be removed. I don't know You've paid like invoices for your company and stuff like this. That's just like ordinary stuff that can be abstracted away on the very, very simple level. And then all the way up to having something where your business can open up into the new customer. And I will say this, like for the customers coming on, they've surprised us. So we have a shift where you run a sauce. you have to understand now that you kind of have to run your business a little bit differently. Search traffic is going down. Maybe people don't subscribe the way they used to.

16:42They want to try your stuff before they pay for it. Or maybe they even require you to pay for usage. And that has historically been pretty hard to do from a payments perspective. So that's part of like a foundational layer right now. Like you have to solve for that. And the new customer, the agent, they kind of require it. They require usage-based, etc. So there's this pressure on providers, SaaS owners, to understand and kind of change into this new way that people want to buy and interact.

17:18Andreas Munk Holm:Michael, I want to ask you, you're heading up everything that's fintech and everything that's payments at Redstone. I'd love to ask you what in the traction numbers made you, both in terms of the entire industry, but also specifically for Solvapay, what made you say this is our right, this is the one we're going with? The honest answer is nothing in the traction numbers, because Solvapay was so early in their journey when we invested. I think they just got started a few months earlier with an idea. Maybe it's interesting to just tell the story of how we got excited about it. So basically, I took the call after one of my colleagues made the introduction.

18:04I had no expectation at all. And during our discussion, I think Vigo saw in my face and my reactions that I thought, wait a minute, that's actually pretty huge. Because there are no payment rails for the agentic world. And the more you think about use cases, the more excited I got. First, I only thought about direct to consumer. Like when I, I don't know, ask JetGPT or whatever solution to buy a plane ticket for me, that's not possible. But if you think business to business and I am talking to procurement agent startups or insurance broker startups and so on. And they all have a transactional component at some point, which at this moment is not solved.

19:00So I thought, well, actually, this is not the incremental stuff I usually see in FinTech. This is massive. So I got so excited that we didn't need to see traction and decided to push the deal forward.

19:17Andreas Munk Holm:Samali, I want to ask you a question here. and I want to focus a bit on the venture investor perspective. Redstone, of course, is investing across the spectrum, but you're not always doing, or in other words, you're not always doing these super early investments. Can you just talk to us a bit about how you as a firm say, okay, we will do it. It makes sense to make an investment that's pre-anything and how you kind of wrap that into the world of Redstone. Oh, it's been an evolution. We see, you know, earlier we were for sure quite a bit more A rounds, but now it's been more competitive in general.

19:57You have to get in early to get a reasonable stake and that's kind of a generic thing happening. The second thing, which is very interesting with this agentic AI kind of companies, is that they very often need only one funding round. If they have a very good business model, they get into the revenue stream very fast and they grow very fast and they never need another funding round again. So if you want to play along and be part of that, then you might have to change this or that at Redstone side or at the venture capital side on your strategy simply to be part of that and being able to be in it at all.

20:33And if you look what's happening, then if they do raise another round, then we're talking about crazy valuations nowadays. So later on, it's pretty impossible for a company in our size to participate later on. So we had to pivot in that sense as well from our side and go with the flow and understand the market, how it has changed. Simple as that. Is the risk higher? For sure. So you just build maybe a bit larger portfolio. you have to very be strict on which one you then support further on and so on and so forth. So it changes a bit of the game as such. But I'm feeling very comfortable with it.

21:10And as Mikael said, sometimes, of course, we don't have all the KPIs. We would love to manage the risk, but that's part of the ball game and going well so far. And, of course, it's about the team. That's a very general answer that probably every early stage VC gives. But it's just true. We look at the team. It's not just Vigo here. It's also his co-founders, all with amazing backgrounds, payments, banks, scale ups, you name it. And then the trust that these guys are going to build something. And in the case of SolvaPay, for me, it was also shared vision that convinced me. I think how in the discussions I had with Vigo during our investment process, how we see the future of payments and how to get there and how aggressively basically you have to build a global company in order to become the payment infrastructure to build like this 100 billion or 1 trillion company.

22:23I think this was what also connected us in a way and gave me confidence to make an investment in a pre-seed company. Also, for sure, it's been quite a change in the last two years. First, we had kind of everything is AI. And I can tell, you know, we don't get any pitch deck anymore, which doesn't say we are the AI, whatever. I'm the only ones who can do it like this. It's a bit crazy. But then to really to figure out where is the really the specialty in there and, you know, combined with the payment, which has been an enormously good business model for a long, long time and having an agentic solution in payment, that was, of course, here especially interesting and also then the good reason to go in early.

23:07So, you know, on the other hand, we also see that Solvapay definitely have a great chance to get into revenues very early on. They don't have to wait for years and build everything for an extremely long time. So very interested about this and very, very curious to see where Vigo brings this.

23:25Andreas Munk Holm:Vigo, I want to ask you, deciding who to raise from, deciding how much to raise, deciding what's the right time, thinking about your future fundraising journey as well. As Sam Lee said, sometimes as an AI company today, you don't even have to raise any subsequent rounds. All you need is the first one. That is true for some. And then there are others where we could definitely see it's the opposite. A company like Grock raised quite a bit of capital as an example. So talk us a bit about how you're thinking about your own capital journey. Yeah, no, great question. I'll say this. I mean, if if there's anything it's just like meeting someone that understands your thesis especially where we were uh where we present a thesis basically and and it's very few talks that i that i sat down into that actually made the connection points and could see we can paint the future in that conversation and i don't have to explain it it was a discovery and and that for me was the signal like okay this this this is aligned this is good because i'm not gonna bash a bunch of other pieces but it's it's super easy to just like try to retrofit and and make what exists and and like paint the world out of that um and it kind of stops there and they're like oh i have to bring in a technical person to explain how this is gonna work and then you know that's not the great fit for us to be honest like you have to be able to see the bigger picture that understand the thesis and we can discuss the thesis and that's what happened with Mikael and Redstone in particular we could skip a lot of the things and just discuss what does this look like what the success look like and what is our like unfair advantage in this I wouldn't be doing this if I didn't do it with Ingmar and Tony.