E654 | Adrian Locher, Merantix Capital: AI Studios & the Future of Venture Building

0:00Welcome back everyone to another episode of the EUVC podcast. Hopefully, by now, the go-to show for European VCs, founders, and LPs who believe in building the future of Europe through entrepreneurship. Today, we zoom in on the intersection of AI and venture building and spotlight a model that I at least often question, primarily for reasons of hubris, but which may now actually be perfectly positioned to thrive in the AI age, at least if it is executed well. Hopefully, doing so is Adrian Locker, co-founder and GP at Mirantics, who have just made a major move by planting a flag in London. Together, we unpack the strategy behind that expansion, the evolving venture studio model, and why AI is fundamentally reshaping both how we build companies and how we fund them.

0:59This show is not investment advice and the hosts of this episode may be invested in the funds and companies featured. Maybe, Adrian, we will get into the big discussion, but just give a quick comment to my statement that it's oftentimes including a lot of hubris when you choose to build a venture model, venture studio model. So first of all, hi, Andreas, and thanks for having me. Well, you know, I think there has been a lot of company building and a lot of studios out there. Some of them worked. I would argue a lot of them didn't really create the outcomes that people were hoping for. What I can say is when I started Mirantix a couple of years back, actually, with my co-founder, Rasmus, we had a model in mind that we ourselves as founders would join.

1:58So in other words, we wanted to build a club that we would join ourselves if we were founders or we are founders, which I still consider myself the most. So I identify more as a founder than I do as an investor, even though obviously the category is I'm an investor. With that much more into it, I have some views. And I do think that it's important to think about what you do when you try and build a venture studio model and really look yourself squarely in the eye. But let's discuss that later. Let's start out with the leap to go to London. Tell me a bit about your expansion to London and why is it the right time for Mirantix to go?

2:37So when we started Mirantix, I lived in the Bay. I was in San Francisco. I'm originally Swiss, so I grew up in Switzerland. But then after exiting my previous company, I moved over to the US just to explore and meet new people and find new interesting topics I'm getting excited about. When we then decided to build Mirantix and get it started, obviously one option was to do it in the Bay, Silicon Valley, or what we then chose to do, let's do it in Europe. And so we had a look at all the ecosystems with one big question in mind. if we want to build machine learning first, AI native companies, we will need to hire the best talent there is.

3:30So it's not about a question, where are we going to find these people, but where will they want to live? And that put to, back then, that's 10 years back, that put two cities on the map, two very cosmopolitan cities in Europe. One was London, one was Berlin. Today, there's a lot more, you could argue, but back then it was mainly London and Berlin. And so we looked at different factors and we ended up choosing Berlin as our home. Now, you could argue that historically and even today in deep tech, one of the most relevant ecosystems is London and surrounding universities you got there. And so for us, it was a very logic next step at some point to also increase our footprint outside Germany, in the UK, in London.

4:32That said, we never considered ourselves as a Berlin or even German venture studio. our ambition was always to identify ideas, validate them, and then build companies that first start out in Europe, but have the potential to go global. Maybe Adrian, just for everyone to be fully aligned or understanding here, Mirantix model, could you just describe it for a second and then say exactly what it is you're doing in London now and the pace with which you plan to take more investment activity there? so morantix is a group of three pillars and we're going to talk about mainly the first one today which is our vc and studio pillar so we ideate and build novel type of companies through our studio model approach and we also invest as a vc in those but also other teams at the precede stage second pillar is what we call ecosystem community that is our campus in berlin that is our campus the london ai hub in the uk and that is also our initiative called ai house during the world economic forum the third pillar is called morantic momentum it's a consulting implementation firm that helps governments and large companies to implement AI machinery projects.

6:02Maybe just talk a little bit about the value of having these three together. So these three pillars at first are standalone independent companies and pillars and teams, also with their respective leadership. But obviously, there are a ton of synergies. When you think about us as an investor looking at new ideas, new teams, I have the whole campus to address when it comes to finding engineers who can help me due diligence, a very specific idea. or I have a very talented team just sitting right next to me, like 10 meters away with momentum, where no matter what type of question I'm going to ask, I'm most likely going to find at least two people with a PhD in exactly that domain.

7:02Then on the commercial side as well, obviously having momentum, having the campus with those massive networks into the European industry, We have a lot of doors we can open for the ventures in our portfolio or even at the stage of ideation where it's about, oh, hey, let's identify the first design partners and let them join very early on, which is usually one of the big challenges when you build a new deep tech company. You need to find commercial partners, identify clients. But how are they going to sign up with you when you're just a two-person early stage company? They won't necessarily trust you.

7:43And that is a bridge we can build because they already know us. They already have interacted with us in one of the areas. And that then allows us to go much, much quicker on the actual validation and also commercial phase. So there's a ton of synergies that come through the group. Let me just ask one question because you have said deep tech many times. And AI is deep tech sometimes. Sometimes it's literally the opposite, which is Vibe coding, and you don't know anything about tech, but you just iterate and iterate very rapidly and get a product out there. Could you just talk a little bit about how you distinguish AI from that second type of Vibe coding AI?

8:29Yeah, no, that's a great question. And the reason why I mentioned deep tech is the type of companies we like to focus on usually have the sort of deep tech complexity when it comes to the business model, but also the technologies behind. So we're not focusing on light wrapper type companies. We are focusing on the application layer, and I think that's an important distinction to make these days. When you say application layer, a lot of people then immediately think of, oh, that's a wrapper company. No, it's not. What we like to build is an application layer that then goes very deep into a respective vertical industry, also integrates themselves along the value chain.

9:24And the business model actually is built on the premise that you're not only delivering technology to your client, but you're actually going to create a lot of value in the whole value chain. And with that, also make it harder for your client to rip you out at some point. That is the type of company we like the most. And to just give you a very concrete example, let's talk about our materials discovery company Cambrium as an example. They build novel types of materials based on proteins. And they have built a protein discovery engine that helps you to identify or even sort of code novel types of materials with very specific attributes, with very specific features.

10:16So they've not only built this engine that they now market to bio or materials companies, but they actually are having a business model that allows them to participate in the discoveries they do. So in other words, they have a part in the IP they create through that platform as well. So whenever a certain amount of material is sold, they're going to profit as well. So it's not just seed-based model. Okay, so now let's go into the venture studio model discussion. You wrote a recent essay where you pushed back against the idea that venture studios are per se. And I could ask you what makes the model especially relevant today.

11:03But in the spirit of having a lot of emerging managers in here, I would love to just before going there, ask you, deciding back in the days when you did to make a studio model, how did you inform yourself around this being the right model for you and it not being, as I called it, a situation with a lot of hubris involved? because that's what I often see when I, the first thing I think is, oh my God, if you think you can build a studio model, you basically have to assume that I'm smarter than everyone else almost if I can keep cranking out these great, great businesses. Comment on that. Yes, I mean, there was, there's probably, it's a combination of two things.

11:48So one is personal preference and the other is my assumption about also financial outcomes. So my personal preference, I've always been a builder. I've been a founder for 20 years. I am a builder. I love building things. I'm the zero to one guy. So even after exiting my company, I was not ready to, let's say, just stand on the sidelines, but I want to be involved. and I also knew from my past experience that there's a lot of stuff that you can help with in the very early days that actually make a difference while there's also a lot of other stuff and now we're talking about a lot of other models who have not worked out very well that just simply doesn't matter so in my eyes it doesn't matter to provide accounting HR marketing and this kind of things to your portfolio companies.

12:52I need founders who can solve that themselves. Otherwise, they're not the right founders for me. But there's a few things that make a difference. In our case, it's true. So we help with validation first commercial traction, finding the right customers and move them over line. And we also help with identifying the right teams and then scale the company quickly. These are two things that in my personal experience also were always the hardest tasks when building a company. So that's where I wanted to support. The second rationale in regards to financial outcomes is my thinking. In early stage investing, it's all about access, finding, identifying the right types of deals you want to make.

13:46And it's also about letting them grow into the company you think it should be. Making these decisions is all about experience and data points. When you do a two-day due diligence and after that you make a decision, most likely it's not going to be as good as if you'd spend two or three months working with the team and then making that decision. That's exactly what we do in the studio model. Yeah. So tell me just a little bit about the inception part, because some studios actually incept the businesses, also including the idea, and others let people come with their ideas and then they work with them for three months.

14:35Talk a bit about where you fall on that. Well, the short answer is we do both. So we do both. We have people who come to us and they'll be like, hey, I'm working on something. and very excited about and we get excited too. And so we decide, let's get it started after validating. Just to make sure we talk about the same things. When I say validating, I mean paying customers. So the secret sauce we have is we usually have paying customers in a company that doesn't even exist for a product that doesn't even exist. So the best outcome that we continue to see when we help to validate these companies very early on, these ideas even I would say, is I'm able to make clients jump at a PowerPoint type presentation where the vision of the product is presented.

15:38and then they join as design clients, as first clients. And then I build the team and then I build the product. So there's no MVP. There's no line of code when the first paying customers are there. That is what we do. That is what we like to validate before even going into the building phase. And that's also in my eyes what makes the biggest difference. Is that every time, Adrian, that is your model, so to say, and that's also the businesses that you look for that can be validated, so to say, on the back of a PowerPoint? I would say mostly. So we focus on B2B type companies, right? Obviously, building a B2C company works very differently.

16:27We don't do that. We try to focus on our strengths. We try to focus on where our assets, where our unfair advantage is, and that is in having all these European industrial network of companies. In some cases, we also identify an issue, a problem, an idea space based on things we see in companies through our network, whether that's momentum or through interactions on the campus, that also hosts a lot of corporates and their teams. We then start to validate these ideas. Other times we see breakthroughs in academia, breakthroughs in science, where all of a sudden two trajectories start to intersect, be it machine learning and biology.

17:23And then we're like, okay, hey, let's dig deeper there. And then we sometimes reach out to people to validate our own hypotheses. And from these outreaches, all of a sudden we find the right founders who want to build this with us. So there isn't a clear modus operandi that we have. I would say these days, it's maybe 60, 70 % of ideas or idea spaces that emerge within our group. And then we start looking out for people who want to build it with us. The rest is when people approach us from outside. I have so many questions. Let me start one here, which is when you don't go out and hunt for talent, I imagine that the recent AI boom has not made it easier to find talent that is ready to jump and who does not cost you a fortune.

18:20or is it the opposite because the opportunity is that much clearer and for that reason more people are ready to go out and build? Honestly, I don't want to sound cocky on this or arrogant, but finding and convincing the best talent really isn't an issue for us. That probably has a lot to do with what we have built already, right? So we have this massive, massive ecosystem with so many extremely talented and smart and nice people, I have to say, that it makes it incredibly easy for us to convince people to join these companies very early on. It's a different war of talent, I would say, honestly, in Europe than there is in the US.

19:11And that has a lot to do, of course, with the level of competition and also the craziness of competition that you currently see in the Bay. That hasn't reached Europe. So no 120 million signing bonuses for your founders? In other words. Okay, let me ask you about the timing. I don't think it's a hard thing to do, by the way. That's maybe another. Let me ask you about the timing question that you commented on in your essay. Why do you think that venture studio model in the AI space is particularly relevant today? AI is a general purpose technology. It is not the cure itself. It's what you build with it.

19:57So the way we understand it, what we are mostly focusing on is what we call the novel types of business models that will emerge on this technology. So it's much less about technology. In those cases, we mostly focus on, but it's about what type of business model can be built on top. And for that, it's very obvious that you need a deep integration into your customers network in the way that you interact with these design partners very, very early on. so you don't build stuff that no one needs. And that is a perfect way you can operate in a venture studio type environment. A second argument I would use is the traditional way of investing early stage was always, okay, here's a team, here's an idea.

21:00And now we find some ways how to figure out whether we believe in the idea in the market than whether we believe in the team. We don't need complete teams and we don't need MVPs to start with. That's usually where in the beginning already a lot of money goes in. If the team has to be complete, if you want to see an MVP, then that requires significant investment. Well, we cut that short. As I mentioned before, we are able to decide to invest and build a company without even having an MVP. and also without having the team perfectly complete yet. Because we also know that the first years in building early stage is pivot land.

21:46So when you think you have the perfect team and then you're going to pivot, well, the team is probably not perfect anymore. And then why did you even hire everybody so early on? Wait until you know exactly what you're going to build and who you need to build it. Yeah. Could we talk a bit about the competition that you have? Because early stage VC is, on the one hand, becoming incredibly competitive. But at the same time, many would also say that it's quite commoditized. Who do you pick? The money, at least, does not really matter. So what's your take on this? Where do you see the competition against your fellow VCs?

22:31Well, I think the change you've seen in the market is probably twofold. One is pre-seed investors became increasingly keen to start investing earlier and earlier, already almost at angel stage when what I just mentioned, there's no perfect team yet, there's no product yet, there's no commercial traction yet. So pre-seed, early stage investors became increasingly comfortable in doing so. And the other one is the seed guys moved downstream. So people doing seed and Series A a couple of years back today show up at pre-seed, which obviously increases competition. Yeah. And why do the best founders pick you in this scenario?

23:32Because we have some great firms that have moved to pre-seed, at least occasionally. It's always a combination of factors, right? There are some opportunities where we have the right to win and where we are a great suited investor. And these are the ones we try to pursue and also to convince. And there's others where we think like, we're not adding the value that then really makes the ultimate difference. And then that's also okay, because I cannot do, I don't have to do 100 investments a year. So I need to pick the five to seven that I think I'm going to be the best to win, but also because I have the biggest lever to make this company successful.

24:20So a lot of times it's a combination of you know this this this commercial network and this access to early clients to also really validate and then also when you do validate so early and then also get the commercial traction quite early on this of course helps you to then raise larger more successful seed rounds much quicker. So in a lot of the seed rounds we've seen in our portfolio, the time from our pre-seed investment to the next seed round was less than 12 months. Why is that so? Because of a lot of traction. Have you ever felt tempted and actually done an investment in a company that you knew you would not be adding that much value to?

25:12so to say they're not going to use your venture studio model at all. They're just on it. And it's about giving the money together with another very strong team. Yes, we do that. Yes, we do that. Sometimes we arrive later, right? When there's already some traction, when there's already a team, there's already first clients, there's already a product, a team raising pre-seed. We're still fine with that. So we can do both, right? We decided last summer to make our model open for both ways of investing. And the bar, of course, for us on the pre-seed side is always quite high because in the studio model, we know how much traction we can get those founders too at the point or at the time of investing.

26:02And so when we look at pre-seeds that are pre-revenue and are very far away from commercial validation, we sometimes get jittery. And we're like, maybe it's too early for us because when we do it through the studio type of approach, we can do so much more and we can de-risk things so much more for the founders and us. But it ultimately always comes down to founders teams. So when I invest, I invest in founders, right? You have been in the AI space from before it was hot and investing. It was cool. Yes, exactly. I'd love to ask you, how have you seen this past period play out? And would you say that it's now easier to be an AI investor?

26:52Would you say that it's harder? Are you more excited now than you were five years ago? or would you just say that now the world around you is more excited, but honestly, you've been equally excited? I think we've always been very excited about, initially we mostly call it machine learning. Today, everybody calls it AI. We're always very excited about what we think is a new category creation technology that enables you to build companies that before just simply wouldn't exist and couldn't exist without this technology. That is the stuff that gets us most excited. What I think has changed over the past couple of years, of course, is the understanding of the broader public of AI has developed much, much further.

27:48I would say still you could argue that most people, when they think of AI, they think of ChatGPT. And obviously, that's the 1 % of what you can do using machine learning and AI. But in general, I think the openness has grown. And also the openness for companies and clients to look into these types of technologies. And with that, also, it's become increasingly easier to convince clients to move early. I think that has changed. That has changed a lot. In some ways also, you see positive impact on regulatory change. So, you know, we do a lot of healthcare. We do a lot of life science and bio. And that's heavily regulated industries, obviously.

28:42And it matters whether regulators are quick or whether regulators are slow. And regulators are always slow when they think it's risky. and when they think there is a lot of potential, they can become quicker. And right now, I think we see a certain tendency for them to become quicker because they also see the potential that technology has. Let's talk a bit about regulation. We recently put out an episode in connection to the launch of the AI Act. I'd love to ask you a bit about where you see the big regulatory hurdles and the big regulatory sweet spots or good trajectories we're on? Well, you know, I think there's a few first principles here, right?

29:29So regulation is something that is required because it sets the frameworks we operate in as a society. And history has shown us that that's the best way to organize society. Now, the debate then is of obviously how much regulation and how much or how complex this should be. The good thing about Europe is, well, we will have regulation. We know what is coming. The bad thing is we don't exactly know what is coming because there's so much still in debate. and honestly also a lot of what looked quite certain a year ago is now again a bit less certain because there's new questions and fundamentally I think the architecture of the AI Act is very tricky.

30:29The AI Act is based on the assumption that you regulate a technology and I think the better idea is you regulate the application. Because technology is moving quick. And when you need to regulate something, you also need to describe it. And when you describe it, then it goes into law. And it's a long process to change law. Most likely, you're going to look at new technologies that don't fall under what you've written in the law three years ago. And so you're also looking at a lot of gray areas all of a sudden that can even be opportunities, right? So I would even argue that in some areas in Europe, probably in a few years from now, we might have opportunities because there's no clear regulation yet.

31:22So people can actually also try a lot. It's really an interesting point in time right now. The majority of regulation is not yet in place. It's such a big topic and you can easily get caught up. The thing is, people sometimes argue that, well, yeah, the US is doing a so much better job on regulation. But then when you look at it in detail, you also realize, well, there's no clear framework in place yet. There is a lot of small pieces around and also even these change relatively quick right now. That is not necessarily a better starting point when you don't know how things will be regulated, right?

32:17And yet, as I mentioned before, I think that argument was very strong 12 months ago. It's less strong now because we also don't know which of the AI Act parts are going to come into force in which exact shape. like you made it pretty clear you think that the most important thing is definitely to have an application focus when you when you legislate rather than or regulate rather than than a technology that's what i believe right so and i'm always happy to be proven wrong but conceptually right now after looking at all this i think that's what what makes most sense but again are you adrian are you yourself involved in some of the work here do or your collaborators in the ecosystem involved?

33:06We are. We've been founders of the German AI Association, which today I think is the largest AI association in Europe, not only operating in Germany anymore, but also on the EU level. So with that organization, where we still are part of the leadership, we do collaborate and we do try to have an influence on things going in the right direction. Do you have any views on the advocacy priorities that you think are most critical for Europe to get right to maintain our competitiveness? Well, I would argue that regulation is one part, but probably bigger levers for being competitive will be making sure we tear down the silos in academia.

34:09because all the great innovation is happening on the fringe of where different types or different dimensions of academia come together, talking of machine learning, biology, chemistry, and all these fields. And in Europe, you still have a tendency of these academic domains being very siloed, right? Another area where I think we need to change how things are going is in which how governments are investing in novel technologies and new ideas. Because today, the DARPA style investing, what I call, is definitely an absolute exception. And the usual type of investing is, well, I'm going to invest in things that don't go wrong, right?

35:07Because otherwise, I can lose my job. And I think that's a huge problem when you think about investing in breakthrough innovation. because in breakthrough innovation, well, most things go wrong, but the few things that work will pull everything out. So that's a typical venture type thinking does not exist there. So yes, we need to be more risk-taking also on these levels. And when it comes to regulation, you know, I think, again, it's important for us as a society to think about what is it we want and what is it we don't. And that then needs to go into regulation. But we also need to make sure we're not trying to think of the next 10 years and how things could look like and then how we want to regulate them when things might come out very different that we're thinking about it right now.