The Techspace Pod, Episode #1: Adrian Locher (Merantix) on Unlocking Europe's AI Advantage

‍

‍

‍Also available on Spotify

‍

In the first episode of the Techspace Pod, we sit down with one of the most influential voices in European AI: Adrian Locher, serial entrepreneur, investor, and co-founder of Merantix. Adrian is joined by our host, Jane Wakefield, former BBC tech journalist, as they dive into Europe’s positioning in the global AI race and what it takes to build real traction in the space.

Adrian isn’t just an investor — he’s a builder who has:

• Launched 10+ companies, including one with a €100M exit
• Co-founded Merantix Capital, backing some of Europe’s most promising AI ventures
• Played a pivotal role in establishing AI hubs across Berlin, Davos, and most recently London (in partnership with Techspace, Tech Nation, Google, AWS, and Husayn Kassai)

With bold clarity, he shares his long-term vision: “To make AI the driving force of our economy."
‍

Top 5 Learnings from the Episode

‍

1. Europe’s AI Advantage Lies in Applications, Not Infrastructure
   “We’re not competing with the US or China on foundational models” But Europe can lead when it comes to applied AI, especially in regulated sectors like healthcare, finance, and energy. Adrian sees the opportunity for Europe in building vertical AI companies that solve real-world problems, rather than chasing the next LLM.
   
   
2. Berlin's "Unfinished Energy" Is a Feature, Not a Bug
   “There’s something valuable about Berlin’s chaos.” It draws in the kind of founder who doesn’t want to plug into a perfect system, but wants to build something entirely new.
   The scrappy, experimental nature of Berlin fuels innovation in ways that more mature ecosystems can’t.
   
   
3. Real AI Companies Must Be Defensible and Useful
   “It’s not enough to have a demo.” The question is: can you build distribution, defensibility, and a meaningful moat? Adrian emphasises the importance of traction and clear market differentiation, especially as the hype cycle continues.
   
   
4. The AI Economy Needs Physical Infrastructure
   “We built the AI Campus in Berlin because these companies need more than capital” Companies need places to collide, collaborate and grow. That same thinking is now driving the AI Hub London, a new initiative from Techspace, Merantix, Tech Nation, Husayn Kassai, Google and AWS.
   
   
5. AI’s Impact on Neuroscience and Education Will Be Profound
   “I’m particularly excited about what AI can do in decoding the brain and personalising how we learn. These aren’t just product innovations — they’re societal shifts.”

‍

Check out the full conversation below

‍

About The Techspace Pod

The Techspace Pod cuts through the noise to spotlight the European founders, operators, and investors building what’s next — from inside the change-making teams shaping tech in Berlin, London, and beyond.Each episode delivers sharp insight from Techspace member companies and the people redefining the ecosystem – from frontier tech and funding, to team culture and scaling strategies.

‍

__________________________________________

‍

Episode #1: Adrian Locher (Merantix) on Unlocking Europe's AI Advantage

‍

Jane Wakefield: Hello, I'm Jane Wakefield, a former BBC technology reporter and your host for this pod. And today I'm delighted to say I'm joined by Adrian Locher, the general partner and co-founder of Merantix Capital. Welcome, Adrian.

‍

Adrian Locher: Thanks for having me.

‍

Jane Wakefield: Now, I wanted to start by asking you about your company. It was founded in 2016. What is it that you specialize in and tell us? Give us a little bit of a flavor of the sorts of things you invest in.

‍

Adrian Locher: So we're all things AI. In 2016, we started to invest in AI companies, helping founders to build their companies in the AI space, in a broader AI space. The way we define it is we call it AI First, which means we're mostly looking at business models that without AI wouldn't even exist.

‍

Jane Wakefield: So let's take you back to the before Merantix days. What were you doing and what led you to want to work in the world of AI?

‍

Adrian Locher: So I'm an economist by training, engineer by heart. I started my first software company right out of uni, and I was lucky enough that this went really well. So we ended up selling the company at some point. Helped building more companies. Had two more exits before then turning investor, and I think at heart I'm still the founder more than I am an investor. But yeah, I started building my first my first companies, when I was, when I was a student.

‍

Jane Wakefield: And what were they?

‍

Adrian Locher: So we started in software right in the early internet days. So in the early 2000s, we started building, you know, websites, content management systems for, for companies. Ending up having bigger and bigger clients. And then I also helped to build an e-commerce company in Switzerland that turned into one of the largest e-commerce players in Switzerland. Then exited that to a media company and also was involved in building a neuroscience company in the US. And yeah, I think I was always I was always kind of like this, you know, entrepreneurial, driven person that just didn't like things as they were but wanted to build them in a new way. And so I guess being a founder gave me that freedom. 

‍

Jane Wakefield: And back in 2016, that must have been a very different prospect to what it is now. So what was a company, an AI company like in 2016 as opposed to 2025?

‍

Adrian Locher: So the very first company we built was Vara in the medical imaging space. They were doing breast cancer screening using AI to automate the screening process end to end. So that means in 80-90% of the cases, you don't even need a doctor involved in the process anymore. That was the first company we started.

‍

Jane Wakefield: And how many do you have now and what what are you really excited about in terms of where AI is going at the moment?

‍

Adrian Locher: So we've got 16 companies in the portfolio and some more to come as we're in the middle of the process to invest in some more. We're mostly well, we're excited about a lot of things. But the way we think about AI is we are looking at an industry and ask ourselves, how much change will AI machine learning bring to this industry, and how many new opportunities will be created that can only exist because AI now is working. And let's look at a few examples. I'm very much excited about the tech bio biotech, a realm in which you have total novel ways of, you know, building proteins, building molecules. And they are so complex that before machine learning, it was almost impossible for us to design. But now, with the help and you could say that in this case, AI is like a tool. You can and you can therefore build totally new structures. So Cambridge and our portfolio as an example is doing bio based materials. And they use machine learning algorithms to simulate billions of potential candidates to then narrow that down to maybe ten, which they then go into wet lab testing and they come up with a with a novel type of material. And the interesting thing here is the impact that machine learning has is instead of four or five years in the lab process, it takes them a couple of weeks to come up with a novel type of material.

‍

Jane Wakefield: Now, here in Berlin, in one of Techspace's beautiful kind of co-working spaces. What is it that Europe brings to the AI party? We see an awful lot of the established big tech companies developing AI. What can European AI companies kind of add to what we're already seeing from those established players?

‍

Adrian Locher: Yeah, first of all, I think when we look at AI as a general purpose technology, then the obvious question is, okay, what industries will be transformed? And the way we think about applying AI in Europe is, well, have a look at all these industries where you have a lot of European companies leading the world's markets, be it pharma, be it material sciences, be it machinery, manufacturing, robotics, even banking and the finance world, there's a lot of leading companies that have grown out of Europe. And we think that the best way to go about is working in these industries where European firms have an edge and think of novel types of business models that will be able to exist in these industries where you have all the ecosystem, you have the existing industry and clients, but you also have the research and you have the talent rather than trying to compete with models that might be better suited for a US company be to be starting in now. That said, when we invest in a company, we don't want it to be in a European company, so we want it to be a company that is starting in Europe and then becomes a global leader.

‍

Jane Wakefield: And your relationship with Techspace is around the AI hub, which is being set up in London, is based in Techspace's office there. Tell me a little bit about why we need these physical spaces for AI, as well as obviously the digital realm in which they will exist.

‍

Adrian Locher: Very, very cool question. And I'll also give you a bit of a background here. How this all started. So when we started building Merantix, we started to invest and build first companies. Right. And I always wanted to have the portfolio companies in one building under one roof so that the teams could collaborate, the teams could have change and could help each other out. And over time, that turned out to be a bit difficult because we grew quite quick. So in about three years, in the first three years after we started, we had seen five different office spaces from the inside because we always had to move as everything was growing. And so at some point in 2019, I said, hey, I want to, I want us to move into a place that will not be too small in 12 months from now, but that somehow grows and breathes with us. And so we were starting to think about, well, we have these amazing synergies in our portfolio companies and the teams. Why not opening it to the wider world? So why not opening up to everyone who was working on machine learning? And with that, the idea of what later would later was the AI campus was born when we found this this place in in in Berlin, Mitte, close to AEG, the former pioneer in in electricity and in Germany. I think it's on 7000 square meters today on three floors where we have 700 desks and 100 different teams and approximately 1500 people working in machine learning. And once we had this place up and running, we were like, okay, so this is now this is now Berlin tech solved. But AI is happening in a lot of other places as well. And we also want it to happen in a lot of other places. So how can we connect these ecosystems with each other? Because building and connecting ecosystems was always part of our mission besides, you know, investing in AI. And that's also how we obviously started to look into London because London has been a hub and is a hub of of deep tech and AI and many other technologies for for very long. And so we started to look into how can we build an AI place in London without actually operating it ourselves because operating this whole building here in Berlin is also quite intense. And there's a lot of operations involved. So we thought about partnering up, and that's how we ended partnering with Techspace and Tech Nation of the Founders Forum Group, Husayn Kassai, AWS, to run the AI Hub in London.

‍

Jane Wakefield: So what does that look like? I mean, what's actually going to be happening in that hub? How do you see that kind of playing out?

‍

Adrian Locher: So in many ways, I think it will be on it already is. So we opened up in February, right? So we already have two months of of of experience and more and more members joining the London place as well. In many ways, I think it will be very similar to what's happening in Berlin. It's a very, very active and large community. So in in in Berlin last year, we've hosted 300 events in total. I think our reception and and security says that it was 100,000 people entering the building throughout the year. And the cool thing

is, with these events, you can see everything from, you know, a large 200 people, panel discussion event with, you know, famous founders, interesting investors and, and everybody popping in and having these interactions to very small 15, 20 people, you know, paper reading groups discussing the latest machine learning papers.

And so a lot of what happens on the campus, today is, is driven by the community themselves. So when you look at these 300 events, I think Marantix, we as a group have hosted 30 events. The rest was organized by the community. And I think London is going to be, very similar in that sense.

‍

Jane: And what do you think will be the outcome of that? Is it that we will see more investment in AI or that companies will learn from each other and share ideas? What, you know, if you were to summarize what you hope to achieve with the AI hub, what would it be?

‍

Adrian: I think all of what you said. And in the end, it's all about the ecosystem, right? Building a strong ecosystem will help all domains, be it the investment realm, be it the research realm where it's all about sharing and developing ideas, testing them out, and then building a company on the base of that idea, but at the same time also involving other groups of stakeholders. Right. So when you look at the Berlin campus, you've got startups, but also corporates who then become early clients of these startups. At the same time, you've got university research groups that are in the building, working on the latest research and then also again, sharing that with the community. You've got investors as well, and you also got government agencies who can also be early clients of these companies. And so the goal really is to find a broader adoption of this technology, because I think sometimes we in our bubble are not aware that even if everybody talks about AI, AI in real life is nowhere, right? So if you enter a random place, be it the hospital nearby or a manufacturing plant in your neighborhoods, you'll find less than 1% AI actually applied. And yet people talk about AI as it would already be there. So in many ways, it sometimes reminds me of the early 2000s and the internet when everybody was talking about the internet, but still there was very little that actually was already economically happening. And so our biggest goal is to make AI a driving force of our economy.

‍

Jane: It's interesting that you make that analogy with the internet because I was thinking similarly that what you describe reminds me of perhaps the early days of Silicon Valley, when they were building the internet and building these businesses that would sort of become the huge companies that we know now. Is that what you're trying to replicate, a kind of Silicon Valley model in the hub? Or is this done very much with a European AI, and you want to do things differently from how things were done in Silicon Valley?

‍

Adrian: Oh, you know, I think there are probably two hearts here in my chest. One is I've also lived in Silicon Valley, and I truly adore this culture and the drive that you've got there. And at the same time, I think it's probably not smart if you try to copy something that has been growing over 70 years now. So I think we need to find our own ways to define things and make them happen. And they should also very much reflect our culture and, you know, our way of doing things. But yes, if it's about the idea of Silicon Valley being an ecosystem and being a breeding ground of new technology disruption and the greater good in mind, the greater good for society in mind, then yes, that's very much what I think about and what I dream about now.

‍

Jane: Something else that you've said, which is very specific to AI, is that Europe shouldn't compete on the application, on the infrastructure level, but on the application level when it comes to AI. Explain what you mean by that.

‍

Adrian: Yeah. So first, we probably also need to define the terms because sometimes people are talking about the application layer as in, yeah, it's just, you know, some software using an LLM so it's sort of like a wrapper. The way we understand the application layer is going way beyond that. And I'm going to dig a bit deeper into that. But the infrastructure layer, as in, you know, the compute, the large models, and then also the tooling layer, I think it's very hard to compete in a world where, you know, maybe there will be four or five large players offering compute and the data centers for AI and the models. So I don't think you will see 10 or 20 companies in this space. And in a way, it's very comparable, I think, with cloud. So you've got four or five large players in the cloud, and then you've got a host of other more specialized, more vertical domain players. But really, it's four or five very large players. And it's an economies of scale game. And I think large language models, foundational models, will be very much the same. And where can you scale a company better? That is very dependent on capital and early adopters. Well, this all speaks more for the US than it does for Europe. So I think the majority of these players on the infrastructure layer will be US-based, no doubt.

When we then think about the more industry-vertical players on the application layer, then I think you will see companies coming up all around the world, based on their individual strengths in their ecosystem. As I mentioned before, I think there are many, many industries in Europe where we have an edge, where we have a lead. And that's also where I think you'll see a lot of interesting companies coming up. Now, when I say application layer, what I mean, or also what we mean by that is we're looking at vertically integrated models. We're looking at novel types of business models that emerge on this technology, where you can then build applications that are end-to-end integrated. And I made this example in the beginning with Vara, the breast cancer screening company. They don't just offer the algorithm, but they offer the screening service. Right. Cambridge, which I also mentioned in the protein engineering space, they don't just offer an engine to come up with all types of proteins, but they very much build and then also sell these materials. So they're much, much more than, you know, just a wrapper. And so that's the way we define the application layer. And in that sense, yes, I think Europe has a great chance in building world-leading application layer companies and will face much more competition on the infrastructure layer.

‍

Jane: And is there something special happening in Berlin? What is it that Berlin is doing differently or doing in a way that other countries aren't?

‍

Adrian: Oh, that's a hard question. So I think, you know, I spent a lot of time in London these days. I spent a lot of time in Switzerland, where there's a lot of ecosystem play and work, you know, great universities, great talent. And I think when we chose Berlin as our first hub back in 2016, I was just moving back from San Francisco. So we were very much looking at London and Berlin, and we then chose Berlin for the first couple of years because we thought that it was a place where it was very easy to attract the best talent from all around the world to make them move to Berlin. And that definitely has proven right. So if you look at our portfolio, there are probably 40-plus nationalities on the teams, and a majority of people have moved to Berlin to work with us, for us, in one of the companies. And I think Berlin has this rough, edgy vibe. It's unfinished. It's not accomplished yet. So I think this is also something that you'll, you'll then find in people who live here. They're willing to take huge risks. They're willing to rebuild things. And, you know, you look at the city that is not polished.

It's it's not a pretty city, right? It's a cool city. And I love living here, but it's not pretty. So, let's let's, let's let's be honest, that, it's got lots of opportunity to do things new because it also is still, sort of like a construction site. And I think that's what makes this Berlin Berlin vibe.

But honestly, like, we're not building Berlin companies. We're not building European companies, right? We're looking out to start something that can then grow global. And again, it really comes down to, you want to build companies where people want to live. So that can be Berlin. That can also be Munich, another great city, with lots of clients.

Also around the corner, because a lot of economic activity is, is going on there and then also be London, Paris or Zurich. So I don't like these comparisons. Well, okay. You know, this is going to be only Berlin and let's do it all here. So I think of of AI as something that happens all around Europe, all around the world.

‍

Jane: Now, you mentioned, risk earlier. And I guess one of the biggest questions for AI companies at the moment is investment and how they get that. How do you kind of make sure that the founder finds the right investment in these kind of very tricky times? And with so much hype around AI at the moment

‍

Adrian: You've seen any hype in the market?

‍

Jane: Oh, just a little bit. Just a tiny.

‍

Adrian: Yeah, obviously. Right now it's it's it's interesting. Right. So sometimes you think that you've seen it all when you were, when you were around in 21, and then, and then you get to remind that that we as humans are probably kind of like built in a way that we like to fall for, for hype.

And we like to be in these in these times. So, yes, I mean, right now, you go out and you see the whole VC market still in a rather difficult spot unless it comes to AI there. It's unconstrained and, and, everything goes and valuations go through the roof. Investment volumes and pre-seed rounds go through the roof, to levels that have not seen before.

Some of it makes sense, some of it doesn't. The way we operate, when we look out for investment opportunities is very much this, you know, business model driven thinking. And it may sound funny because we're a tech investor that we spend majority of the time thinking around the business model, because that's also what, in the end, will create the value or not.

So the question is, is what you're doing really adding a ton of value to your client base? And will you be able to capture that value in a, in a in a smart business model that also is able to build up moat over time? Because building moat on the pure technological side is quite hard in a technology that is commoditized, very quick, and is also changing and evolving so quick.

So we really look into the real world adoption of this technology. And when we look into invest in cases, we want to see a lot of client traction. We want to see first design partners interacting, with an idea and the earlier these design partners interact, even at the pre product stage when there's just a PowerPoint, and they sign up to be the first clients, the more you know that they are actually there is actually something they really want to be there.

And they are willing to, you know, take that huge risk of partnering with a two people startup at the at the pre product stage. So there must be something that they're seeing. And there must be some value that, that, that they can create. So we really approach it from that perspective rather than, you know, just, oh, how much different is this technology.

‍

Jane: Now we're seeing AI being applied to all kinds of industries, from medicine to law to finance. Is that an area that you think has yet been untouched, that you would like to see AI be used in? You know, someone come to you with an idea and you're like, yes, I want to invest in that because that's new.

‍

Adrian: Yeah, I mean, I mentioned robotics, finance, bio before. There's obviously a lot going on already. I think two areas where I'm also very, very excited for different reasons, that are much less developed yet are education and neuroscience. So in education, I think, the obvious case, but something that is not really out there yet is, you know, when you look at what's the best way of educating someone, you end up looking at tutoring, one-on-one tutoring. Yet that's not scalable. So that's why most education systems cannot be built in that way. But I think for the first time, we are looking at technology that will enable one-on-one tutoring in a scalable way in addition to, you know, in-person interactions in education. I think that's going to be a huge shift and it's going to be a huge change and will also democratize education even further. That's why I'm very excited about this whole space yet. From a business model perspective, education oftentimes also is very difficult. It's highly regulated. There's a lot of countries where, you know, governments have monopolies. So it's harder to break in. It's harder to disrupt. But I'm very sure it will happen, and it should happen for the greater good in education.

Jane: Yeah, we definitely need some radical change in education, I would say. I used to be a teacher and it feels like it hasn't changed for many, many, many, many years. And you mentioned neuroscience. So we were chatting off-camera about having chips implanted in our brains. Well, what do you think AI is going to bring? What are you excited by AI bringing in neuroscience?

‍

Adrian: So, look, I think when you look into the medical progress that we have or we've done when it comes to the brain, it's actually quite limited. So the understanding we have about the brain is still in its first innings, I would say. And, that said, I'm not a specialist in neuroscience. Right. So, just the way I look at it and the way I interact with many people from that domain, there are a lot more unresolved, unanswered questions than there are answered questions. And one of the reasons is the brain is a very complex organ that is very hard to grasp. And when you think about how we as humans usually understand complex systems, is a lot about simulation. And I think machine learning, for the first time in human history, gives us the tools to start to simulate brain activity at scale, and that will ultimately lead to a much, much better understanding of what's going on and when we do so. I think there's two major developments that will happen. One is, well, a lot of the central neural disease-based diseases will be much closer to being solved than they are now. So in a way, you can think of it as, you know, the moment we were able to decode the genome and what kind of impact that had on cancer and now and treatments based on cell on cell systems, I think there's also going to be that moment in the brain. And a lot of these brain-based or central neural system diseases, they are now starting to haunt us because they're also tied to demography. Think of Alzheimer's. And a lot of people think that as an example, Alzheimer's ad will one of the most costly diseases for health care systems in the near future. So we better get there and solve this quick. And I think machine learning will be a big driver. AI will have a big impact on also understanding these complex diseases by understanding the brain. The other way, of course, is or the other early, of course, is when we understand the brain better, we'll also be able to build even better, smarter AI systems. Right? Because a lot of the current best practice models are somewhat based on structures that you find in the brains and methods in how the brain operates. And the better we understand, the better these systems most likely will get.

‍

Jane: It's certainly going to be an exciting journey. Now, I've thrown loads of questions at you. We want to finish by getting you to think of a question, you know, what's the burning issue that you think we need to think about? It could be in AI. It could be just something related to Berlin. A question to finish with.

‍

Adrian: Yeah. I think, you know, when when I, when I think about going outside of of of of our bubble, of my bubble, I want AI to become something that anyone can use and, something that will elevate us as us humans.

And one question I keep pondering is, you know, if we would be more conscious, in our, in our behaving, in our being, would we be better able to create a better world? And I somehow have the hope that I will help us to get there by elevating our own cognitive system, our own cognitive behavior. Maybe by taking off some load, on stuff that we don't need to process.

So we have more free space in the brain to process other, but then also by just becoming more aware of, of what's going on, and, and in the US, in the US, being able to create, you know, the next version of the world we would want to live in. It's a bit of a philosophical one, but that's, what I really think about.

‍

Jane: That was a very good question. I'm not going to attempt to answer it myself, but I'm sure somebody will someday along the line with a bit of AI help, perhaps. It's been absolutely fascinating talking to you. Thank you so much for joining us.

‍

Adrian: Thank you. Thanks for having me today.

‍