How I Invest with David Weisburd - E411: How AI Is Creating a New Generation of Venture Firms

Episode Date: August 3, 2026

Most venture firms are using AI to save time. Earlybird is using it to generate alpha. David sits down with Andre Retterath, General Partner at Earlybird, to discuss how his firm built an AI-native venture platform, why proprietary data is becoming venture capital's biggest competitive advantage, how machine learning improves investment decisions, and why the future of venture belongs to investors who combine technology with exceptional judgment.

Andre, you joined Early Bird while you were still completing your PhD and machine learning. What made you join Early Bird? To be very honest, I just Googled what's the best venture capital firm in Europe. And I looked through their portfolios. I looked through the teams and Early Bird to me was the most technical one. So we already had back then about 80% of B2B companies in the portfolio and being an engineer myself. I was looking for like-minded people. And the founders of Early Bird also had a PhD, an aerospace engineering, industrial engineer.

and so on. So to me, that felt like the right place, and I just sent one application and got in. How did your PhD in machine learning? How does that apply to venture capital? Everyone said you cannot automate venture capital. It's a people business. It's all qualitative data, specifically in the early stage investing. And I thought, I disagree. And that triggered me to do some research. So initially I spoke to industry experts. And then later on, I just started building stuff that helped me also throughout almost the past decade to really automate lots of the redundant, repetitive work in venture capital. I want to get into in a bit how Early Bird's this AI native platform.

But before we get into that, I'm curious when you joined Early Bird and you joined the venture capital space, what surprised you most as an outsider? It's a very, very close industry. And to my surprise, at least here in Europe, most people knew each other. So they go to the same schools and I think it's mostly the same also in the US and other ecosystems. But I was very distant to these networks. So I didn't know anyone when I got into the industry and I was surprised how much and how close these networks were with the existing investors in the industry. So it was quite difficult to break out from the outside, not having been at one of the top business schools whatsoever.

Do you think that's related the closeness of the industry with the group? think? I do think so. Venture capital as an industry got started in 1950s to really have independent thinkers that fund the most risky businesses that would not get capital elsewhere. So those businesses that would not get funded by traditional banks, this is where venture capital initially originated. And I think over the past whatever 70 plus years, the industry has become quite much of a group thing. I think that's also driven because the industry is very, much of a power law distribution. So we all know the peri to principle 8020. And venture capital has a very high alpha coefficient, meaning there's a higher concentration even. So about 5% of your

portfolio eventually drive 95% of the outcomes. And that being said, everyone knows you cannot afford to miss these outliers. That means false negatives are suddenly very expensive, saying no to the wrong opportunity becomes very expensive. And that being said, there is this former of investors and they're always looking around to not miss this one thing. And if they hear that your friend is looking into this one opportunity, they suddenly feel, oh, if they are looking into this or if Fund ABC is looking into this, I should also look into this. And I think the combination of this close network, everyone knows each other, everyone is talking, the gossip kind of world, the awareness of the power law, the concentration in the industry, and then also the high cost, like these adverse costs

of the false negatives, that the nose are very expensive. The combination of the three really leads to investors looking into the mainstream and getting into this group thing. When people talk about portfolio construction, they talk about the math, but not necessarily the philosophy. How would you explain early birds' philosophy when it comes to portfolio construction? Our philosophy, I would describe as very disciplined. We had in the past always been elite and business.

So early bird next year turns 30 years old. So it's one of the oldest, one of the largest and most established funds here in Europe. And we have been in a lucky situation where in the past, the market was very much supply side constraint. Supply and a sense of capital and demand and a sense of startups raising money. So when I joined early bird for the first time in 2017, it was a very interesting experience. We had one investment committee per month. We flew in the different teams to this investment committee and we have met four teams subsequently throughout the day. And if you met a new opportunity, the Thursday that after, you could tell them like, we need to wait another three and a half weeks and we could afford to do this.

And we could also afford to just lead rounds completely ourselves. There were no followers and around, no angel investors. And we got very strong shareholding. So for some of the companies that are also mentioned, your iPod is one of these examples. that really falls into the history of early birth, where we invested relatively small amounts, but got very strong shareholding. And if I'm talking about strong,

I'm talking about towards 20%, sometimes even more of that for initial shareholding. To your question on how I would describe our strategy, we had been very, very disciplined, and thankfully also had been disciplined throughout the COVID period. So we rejected lots of opportunities based on price or because we couldn't see the sustainability of the business model, which today we are very happy about.

but we also missed some opportunities because we rejected you to price or shareholding. So for example, I myself, I rejected lovable. We looked at it at the pre-seat round and we rejected it because we would have been only at 8% or something and that was below our threshold. So for the fund, we always aim for 15% initial shareholding on average across the fund. How costly was that mistake in lovable? It is very expensive and that's the problem that I mentioned before because we have this asymmetric cost matrix where if you lose your money, you can only lose it once.

But if you do not invest, for whatever reason, in this example I mentioned, we did not invest because the shareholding didn't meet our criteria. Then you will lose multiple times the upside. So on a lovable, I would need to do the exact math, but I would not be surprised if it would have been north of 50x and most likely a fund return a level investment. And I can give you more of these examples where we reject it oftentimes because of shareholding, sometimes because it was too expensive, sometimes because the round dynamics did not match whatsoever. So we all have these anti-portfolios and I think they are very costly,

but we need to learn from our mistakes and then revisit our portfolio construction and really revisit the strategy. So as a consequence of that, we have also selectively become more flexible on shareholding, for example. This anti-portfolio, I think Bessemer really popularized that they still post this anti-portfolio of all the companies that they missed. One of the things that a lot of top VCs tell me is that there are companies that end up doing the very best at the local level, at the round level, are always, quote, quote, overpriced. Do you find this correlation between rounds that are overpriced at the local

that end up being these 50x returners like lovable? There's actually data on this, and it shows that companies who raise the early rounds at higher valuations and at less dilution have a higher likelihood to raise subsequent rounds in bigger amounts and lower dilutions. Meaning, if you want to reframe, founders who are great at raising continue to do so because it is a skill to create this momentum. And I think there are some examples, for example, in DeepTech. DeepTech is a great example. Very capital intense. You need hundreds of millions, if not billions, for many of these business models. The same is true for these AI labs. They need hundreds of millions and billions

for compute. And I think one of the capabilities we look in for in the founders is really the ability to attract capital. Is the CEO able to attract capital? And I think this is really a key criteria that we want to see in these capital intensive businesses. It's interesting because in startups and investing oftentimes the hard skills are. Everybody agrees there's hard skills, there's great engineers, but people like to pretend like fundraising and recruiting these softer skills are not actual skills and don't really drive the needle. In many cases, and depending on the industries, that could be the thing.

In hard tech, it's much more about can you attract the talent, can you attract the capital, than are you the world's best engineer? 100%. We see this skill become more and more important. So it's really be more of a coordinator. So also in a world of AI where you need to coordinate different agents. It's less about being an expert in all of the disciplines that matter. As a CEO specifically, it's most important that you can bring together the right people,

the right stakeholders behind your mission, your vision. That's number one. You need to attract the right amount of capital to really implement and really be set up to win on a global scale and not just on a local scale. And then also it's very important to prioritize over time. So let yourself not distract and go into one or the other direction. but really go full force ahead on the right direction and really align all of the people behind that. And I think, as you mentioned, some would call this a soft skill.

So I think these skills for CEOs have become incredibly important. Also, founder branding. Lots of people are thinking, why should I put myself out there on LinkedIn, on X, whatever? But for many of the companies, it has become the most attractive channel for hiring talent, for partners, for customers. So I think some of these soft skills will become incredibly important in an era of AI. When we talk about soft skills, the term that I use, my own terminology, I'm reminded. I have Alex Hermosie in my ear saying, well, soft skills are just hard skills that people have not found a way to quantify.

So they're not really soft skills. There's just hard skills that's just harder to put your finger on. 100%. Yeah. No, I love this. And I think it really depends. So for us, we think quite a bit of like what are different frameworks. So we can make sure we all are on the same page if we evaluate investment opportunities.

But for us, we use a simple kind of framework. We have deep tech, deep tech for us really means hard tech. You can touch it. So for example, we've invested in an ice aerospace, which is a lower orbit rocket launcher or Marvel Fusion. This is core fusion or green light, which is decarbonization or Arago, which is a new chip infrastructure. So this is really everything that you can touch. What sits on top is the software infrastructure and frontier AI models.

That's also the investment practice that I'm leading at Early Bird. And there we invest across all of these foundation models and everything that really enables them from the data collection, data infrastructure and so on and so forth. And then at the highest level, we have the application layer. And at the application layer, it's really about taking existing components, assembling them and really building a solution that's either for one specific department. so for sales, for marketing, for HR within an organization, or goes after a very specific sector. So construction tech, industrial tech, e-commerce, whatsoever. And we have split our team in that way. So we have a deep tech sub-team.

We have a software info and AI Frontier sub-team. And we have an AI application sub-team. And these teams look for very different criteria in the founders. So for example, if we look in deep tech, deep tech, we want to see people who have a very strong research foundation. They have IP, they have published papers, they have the credibility and really know also what it means to bring research into production. If you want to produce something that's hard, it also means building production facilities, understanding process management, supply chains. So all of this stuff is very, very important. If we look into the AI Frontier models, we really want to also see the CEO specifically, someone have the credibility that you can attract the best research talent out there.

So if you build, for example, a world model company or a voice infrastructure or an image or video model company, it's like whatever, say two dozens of exceptional people out there in the world and they want to work with the very best. And we want to see in the team, do you have the gravity to attract these best people versus in the application layer? We look for people who understand the respective industry, the respective department, because you're selling into those and you need to understand the problem. So across the spectrum, we really look for very different criteria. area and the founding teams. It's so interesting because a lot of times venture capitalists talk about these founder traits, these canonical founder traits as if they're one size fits all, but every industry just has such idiosyncratic qualities to that a founder could be very good in one industry

and just be a terrible founder and another one. I couldn't agree more. I'm still thinking about this fact that you mentioned, which is when you do the research, the rounds that had the lowest dilution at the highest valuation early on, continue to do this. So this fundraising was the skill, which makes sense. But I'm thinking, what are the second order effects of a great fundraiser? Are there other advantages that accrue to that founder? If you look at the underlying characteristics, it's that someone can sell you a vision. I think fundraising, and there are lots of examples where if you look at the fundamentals and the

health of the business, lots of people would likely not invest. But the founder has the capability to convince the shareholders. So Elon Musk is a great example for this. He could convince people to join the Tesla, the Space X journey, and we've recently seen also on the IPO that he was able specifically in the breadth of the market retail investors to convince them. And I think the same trade you can see in founders. If you look at the underlying trade, this also is applicable to talent. Because if you ask people, why did you join this company? They most often have one single reason, and it's the CEO, or it's a specific CTO, for example. I have joined because once in the life I wanted to do research with this specific person.

And I think it really is the craft of selling, but not overselling, because it might very quickly hunt you back if you're overselling and people figure out this too big of a gap towards the reality. So I think this is a skill that you can apply from fundraising to talent attraction, to customer and partner attraction, but it's very important to have this. And for us, it's most important to have this in the CEO. For the CTO, for example, I want to see a completely different profile. There are different profiles of CTOs.

One is like the charismatic, CTO, the visionary who's on stage. There is the individual contributor coder, for example, and there are lots of other CTO profiles. So for us, if we look at the team, we want to see a set of capabilities that are well distributed and fit their respective position. I think one of the most underrated aspects of Elon Musk, obviously one of the smartest people in the world, but his ability to speak clearly and effectively to billions of people. He obviously has a huge Twitter following. And that's such a rare case to be able to be so technically in the weeds,

but also be able to communicate on such a simple and wide-ranging audience. 100%. It's one of the key capabilities. And I mean, if I look back at my research, I focus on two things. One is really how can I leverage machine learning, AI, to automate venture capital, to really look into the efficiency part and also the effectiveness part, meaning with limited input, how can I create more output and process more? But then the effectiveness part also meaning how can I create alpha, meaning how can I see

the deals that other funds do not see? How can I make better decisions? how can I connect the dots in a different way that most people would. So this is one area of my research and the second area was really the data-driven part where I tried to zoom out and look at the data about successful companies. And we've published also the summer of the research through my newsletter, but also through different papers out there where we looked at tens of thousands of companies. We looked at them at T1, meaning, for example, 2015.

We take all of the input data from crunch base, pitchball, harmonic, whatever, whatever databases were available back then. And then we track these companies on, let's say, 2020. And we call this T2. And as of T2, we define what is success. Means company did an IPO. Company got sold for more than X. Company has raised more than Y money and so on and so forth.

So we can define what success looks like from our perspective. And then what we do is we just classify them binary as all of the success case. is R1, all of the failures are zero. And then we look for all of the ones and try to find the patterns as of T1 in 2015. So the successful companies, did they have patterns, that they have something in common five years ago? And this is what we tried to do from research. And there's lots of clarity.

You can look at age distribution. You can look at specific universities. You can look at specific degrees. you can look at social media following and so on and so forth. And depending on which industry they are building in, you can see very clear patterns. And we try to incorporate that into our systems because I, Andre, as an individual, do have specific biases. So I did my PhD at Technical University of Munich.

So I might like students or graduates from Technical University in Munich more. I might have, let's say, similarity bias. I might have recency bias. And all of these biases I can codify, and I just repetitively do every single day what my biases tell me. So I look through a pitch deck, one or two minutes, if we believe the stats, then I try to act on my biases. And in order to find a more scalable way, we looked into the research.

And this is what I described earlier. If we look at the whole universe and the patterns for these successful companies, try to incorporate them and automate the selection. What surprised you the most from your findings? That there are some patterns that do not change over time. So, for example, you can see that age has been quite robust over time. You can see that the average founder is 34 years old, and then you can see a specific distribution on that.

So this age is for Unicorn founders back then, and it has been quite robust for the past 15 years. but then you can also see other characteristics that are completely changing with the respective time we are in. So for example, I mentioned earlier founder branding or personal branding for the team. Five years back, it was not the topic. Today, if we look at the best companies, if we look at the lovable, if we look at the Black Forest Labs, if we look at the spacious, if we look at the whatever companies, 11 labs, Sintesias and so on of the world out there, the founders do have strong brands.

And we can see that they leverage this to also position themselves in a world that has become way more competitive and more difficult to differentiate. There are other characteristics. So, for example, if you look back 15 years, and there was just the penetration of cloud happening, this was a completely different skill set and you were looking for completely different people.

If you look into the age of mobile, the companies that became successful in the age of mobile had completely different characteristics and also different kind of founder backgrounds. So there are some things like age, for example, or also specific university backgrounds that are very consistent over time. And then there are some things that completely change

depending on the error you're building a company in. So a lot is also about timing. Are you the right profile to build a successful company at this time of the market? Most investors don't lose because they lack information. They lose because they can't separate signal from noise. Every day, thousands of earnings calls, SAC filings, expert interviews, and market updates, compete for your attention.

The challenge is in finding more research. It's finding the few insights that actually change your investment decisions. That's exactly what I unpack in my conversation with Ryan Fennerty. Instead of talking about AI in theory, we discuss how leading investors are using it to surface differentiate research, move faster than competitors, and make more informed investment decisions. If your investor looking to gain an edge through AI-enabled insights, I think you'll find this conversation valuable.

You can access this limited release episode by going to alpha-dash-sense.com slash how I invest. That's alpha-sense.com slash how I invest. So it's not just founder industry fit. So hard tech versus application layer. It's also founder industry time and market fit. And maybe the memetic desires of venture capitalists. Five years ago, venture capitalists were obsessing about the 40-year-old founder.

Today, they're obsessing about the 22-year-old founder. And if you get the wrong time frame, you just might be out of luck. 100%. So best example, five years ago, if you would have built a company in Europe in defense tech, you would have gotten zero money. Majority of the funds are not even or were not even allowed to invest in defense tech back then. Now, Germany, whatever, 18 months ago, decided to invest 100 billion into defense. And with that, you can also see the NATO requirements and so on and so forth. Suddenly, defense is a big thing. And if you're a founder and defense in Europe right now, you have access to tons of money. So same founder, five years difference. Five years ago,

very difficult to build a successful company. Today, almost impossible to avoid if you bring the right criteria to build a successful company. These biases are hard to intuitively see because you can't prove the counterfactual. What if Palmer Lucky started Anderl 10 years ago? No one would know about Anderl. He still obviously had a previous exit. Some people would know Palmer Lucky, but he wouldn't be the same person

versus now today where he was able to raise funding. Exactly. And that's really the point. The best founders have kind of a sense about there is something moving or there is like a breakthrough in technology. For example, we invested in 2021 in a company called Alif Alpha.

So I've been reading back then the Transformer paper in 2017. I got into the community 2018, 19, and it was very few people, specifically in the investor landscape that didn't even know what, like most people didn't even know what was going on, frankly speaking. And we at Early Bird built out the thesis and said, look, Andre, you will join Early Bird to build out the AI investment practice. And we've done quite a few investment in general machine learning, deep learning aspects. But then we invested in Alaf Alpha in 2021 because they were the first year. company to actually train their own large language model. And that was ahead of its time. We had investors back then who said,

can you show me like a chat board or something? I don't even know like what is a large language model. We were debating how much of a product should be built to show it to investors because most people did not understand. And suddenly chat QBT happens. And three months later, everyone knows what a large language model is. And suddenly there is tons of money available. So I think there is a lot also about timing.

But what the founder of Alafair has done really well is the sense of we are close to a breakthrough. I don't know if it will happen this year or if it will happen next year, but I'm sure it will happen in the next three years and not in 10 years. You mentioned today founder brands are much more important than they were five years ago. How do you explain that? And what are the second order effects of having a great founder brand? For most founders, it's not intuitive to show their faiths, to be online every single day. because most of them just want to be heads down building a company.

But what most founders forget is that it's a very scalable way. And I mentioned earlier, most often the shareholders, the team, the customers, the early partners, they join a company because of the ability to sell a vision from the CEO, from the founders. That's in most cases. And if you look into the scalability of this ability to sell, this is what I see as a founder brand. So a founder brand for me is a scalable way of packaging your skill to bring people behind your vision and scale it into the world, make it accessible for everyone so that everyone hears about your brand. Everyone can see if they are a good fit and they buy into your story and want to join your journey in whatever capacity. So I think it's for many companies unavoidable.

I think some other companies, if you build a very deep tech, deep in the stack and hard tech, you might have other. other problems for now, you might already have access to the talent. But for most companies that sit higher in the stack, I think it's unavoidable because also the higher you sit in the stack, the more difficult it is to differentiate. Today, when we can vibe code something over the weekend, there are thousand companies that are lookalikes and they tackle the same problem. So how can you actually stand out? If you look into bytecoding, I mentioned this example before, there were probably at a time

40 or 50 companies with similar approaches. But some of them have managed to attract the capital, but also the talent. And I think founder branding was a very strong instrument for that. Out of those two traits, attracting capital and attracting talent, which one's more important today? It depends which part of the stack you're building in. The deeper you sit in the stack, the more I lean towards the ability to attract capital. the higher you sit in the stack, the less relevant the capital is, and the more important, the talent becomes relative.

I think always both is incredibly important, but if you cannot have both, I would say this is a framework to look at it, because if you sit deep in the capital, you can have deep in the stack, you can have the best talent in the world, but if you do not have the capital, you will never succeed. And the other way around, if you sit very high in the stack and you have all of the capital in the world, but you don't have the right talent, you won't be successful. If you sit lower in the stack, you have all of the capital in the world, but you don't have the right talent. I think there's a good chance you can still outcompete some of the peers because access to capital is incredibly important. So the deeper in the stack, if you're power, chips, large language models, the more

capital you need. And the higher in the stack in terms of the application layer, you need more talent. Why is that? Because the higher in the stack, the less relevant the capital becomes. Today, you can vibe code something over a weekend and you can find value proposition market fit right away. So you can just send it to your ex-followers or something and you will have to first test us and you can even bootstrap such a company easily into a few millions of ARR profitable. But then later on only it becomes important to raise more capital and also get the distribution right. So the higher you sit in a stack, the less capital do you need to achieve specific milestones.

The lower you sit in a stack, the more capital you need early on to even show a proof of concept and de-risk so that you can bring it into production. For example, if you build a rocket or something, you need to first build the engine. Then with the engine, you need to show a hot fire test. need to burn for 30 seconds or something. Before you get this, you won't get access to more capital. So yet there are technical milestones which all require huge amounts of capital. And this is why I'm saying the deeper you sit in the stack, the more it matters to have

skills to attract capital. When we last chatted, you mentioned that early bird is one of the first firms really building an AI native platform. What does that mean exactly? When I joined early bird in 2017, I had more than five years back. background in industry, I did process automation, so I was really turned for efficiency and then got into venture capital and saw the most brilliant people I have ever met doing literally monkey work every single day.

So I remember to give you an anecdote, one of the founding partners asked me, look, I will attend this conference in three weeks. Here's a list of all of the attendance, 400 people. Let me know who to meet. And I was like, okay, how should I do this? Yeah, just Google the people. So I started going through Crunchbase back then. I looked up there in Germany.

There was like a version of LinkedIn called Xing. So I looked up different people. I looked up people on LinkedIn. And then I created a scoring. I looked into our basic CRM system back then and said, have you been in touch? Yes, no. That was mostly the only information that was available.

And then after about a week of work, I could tell him, these are the most promising companies. And I was like, I cannot believe that such smart people, around here, spent a whole day doing stupid stuff like that. This can be easily automated. Even stuff like, can we do an introduction to one person? Suddenly, everyone is chatting in the Slack or WhatsApp. Do we know this person? So investors spent so much time with manual and inefficient work that I took this as a trigger and really think venture capital from first principles. And I presented that to the founders of Early Bird back then. And they got really, really,

excited. So they said, look, this is the future of venture capital. We need to invest in this direction. So I joined as an investment professional, which I think was also very important that I did not join as an engineer, for example, because you need to understand the processes end to end. You need to be in the full depth of the decision making to actually leverage technology to change the processes. It's very difficult outside in. If I would need to understand your processes and I would want to transform them, it's very difficult outside. I didn't. It is possible. I could shadow you for, say, two months and then start automating step by step. But if you build something for yourself and you have the skills to actually do this,

I think that's the best setup. And then I started building tools for myself. So simple automations in the first beginning, then knowledge graphs to represent our network, to calculate the network density, like which people do you know best. Is there a new opportunity evolving in the network and so on and so forth. And we scaled it from there. So I built for myself between 2017 and 2020. Then we hired the first full-time engineer because the strategy was set. We knew the different components that we required. And then we'd ramped up the engineering team. So at the peak, we had eight full-time engineers at early bird. So that was about 20% of our total team size, which was quite big. To do the fundamental ground work, the data plumbing, as I call it,

call it. And today we have three senior engineers who really spend most of their time sitting on top of this data infrastructure and really making sure that it is actionable in all of the functions across the investment firm. So we have rethought from first principles what the tech stack of an investment firm should look like. And then secondly, we have transformed and are still in the process of transforming, frankly speaking, all of the processes and workflows, because investors are very reluctant to change their workflow. So a lot is also about change management that we are facing in the last, say, 12 months, as many more of these tools have become available.

I want to get to change management in a bit. Today's Friday. Give me a couple examples of how you're using AI just this week. Look, everywhere. I can tell you simply said, like our stack has changed a lot. So back then we built all of this data infrastructure and we built a dashboard on top. But a bit more than one and a half years ago, I questioned if the dashboard is actually the future way of interacting with data.

So you could see that there were different LLMs and AI chatbots available. And I personally believed already back then and even more so today that the single interface to interact with different software tools and databases, will be AI chatbots. So today we use cloud code across the organization. Everyone uses it. And then it's important to connect all of the data, all of the knowledge that we have through different MCPs with our system.

And with that, I do everything. Like, for example, earlier this week, one of my portfolio companies received the term sheet. So I just throw it in. It benchmarks against some of our existing term sheets in the database. If I'm preparing for a meeting, so I have a sequence of different founders. meetings every week. And I always have 15 minutes in between. And 15 minutes before the next founder meeting, I get a digest, which tells me this is the next company. Here are our interactions. It's a brief summary about what the business does. Core questions, automate a generation of

competitive landscape. We can see all of the traction that is available, recent media, news mentions, and so on and so forth. So you get like a one-page of digest. And if I get into a founder meeting today and probably in 95% of the cases better prepared than some of the other investors who are oftentimes not preparing for some of these first intro meetings. Another example is that if we create an investment memo, for example, I remember back then it was a working document that easily took us one or two weeks to prepare. Today we take all of our existing knowledge and we just say create an investment memo. We have trained that on 1,000 plus investment memos from the past, so I can easily generate it.

Another example, we have built an internal skill. So, like, everyone in the organization builds skills. We have a process where we can submit them. They are reviewed internally, if they are secure, if they are compliant, and so on. And then they get rolled out across the organization. We have one that has the early bird CI in a presentation. So, like, I could take a transcript of our call here, and I could just take the transcript. script and say, I want to present this in 30 minutes at the keynote.

And if we're growing up, I thought managing money meant paying bills and balancing a checkbook. But as you know, that is only a small piece of the financial puzzle. Managing your money takes more than just checking your bank account every once in a while. And great financial decisions come from having a complete picture and proactive management of your income, expenses, and investments. Take control of your finances with Monarch. It brings together all of your accounts, investments, saving goals, and spending into one place, making it much easier to understand where your money is going and whether you're actually on track

to achieve your financial goals. What I like most about Monarch is that it doesn't just tell me what already happened. It helps me plan ahead. The AI assistant lets me ask questions about my finances and plain English. And the AI weekly recap highlights spending changes or upcoming expenses before they become surprises. It's like having a financial advisor in your pocket. Write your own money story with Monarch. Use co-invest at Monarch.com to get your first year of Monarch Cor half off at just $50. That's 50% off your first year at monarch.com with code invest. Support for today's episode comes from Square. The all in one way for business owners to take payments, book appointments,

manned staff, and keep everything running in one place. Whether you're selling lattes, cutting hair, running a boutique, or managing a service business, Square helps you run your business without running yourself into the ground. I was actually thinking about this other day when I stopped by a local cafe here. They use Square and everything just works. Check out as fast, receipts are instant. Sometimes I even get loyalty rewards automatically.

There's something about businesses that use Square. They just feel more put together. The experience is smoother for them, and it's smoother for me as a customer. Square makes it easy to sell wherever your customers are. In store, online, on your phone, or even at pop-ups, and everything stay synced in real-time. You could track sales, manage inventory, book appointments, and see reports instantly, whether you're in your shop or on the go. And when you make a sale, you don't have to wait days to get paid.

Square gives you fast access to your earnings through Square checking. They also have built-in tools like loyalty and marketing, so your best customers keep coming back. And right now, you can get up to $200, awesome. Square hardware when you sign up at square.com slash go slash how I invest. That's SQUA-R-E.com slash go slash how I invest. With Square, you get all the tools to run your business with none of the contracts or complexity. Run your business smarter or Square. Get started today. Support for today's episode comes from Square. The all-in-one way for business owners to take payments, book appointments, manned staff, and keep everything running in one place. Whether you're selling lattes, cutting hair,

running a boutique, or managing a service business, Square helps you run your business without running yourself into the ground. I was actually thinking about this other day when I stopped by a local cafe here. They use Square and everything just works. Check out as fast, receipts are instant, sometimes I even get loyalty rewards automatically. There's something about businesses that use Square. They just feel more put together. The experience is smoother for them and it's smoother for me as a customer. Square makes it easy to sell wherever your customers are. In store, online, on your phone, or even at pop-ups and everything stay synced in real-time. You could track sales, manage inventory, book appointments, and see reports instantly whether you're in

your shop or on the go. And when you make a sale, you don't have to wait days to get paid. Square gives you fast access to your earnings through Square checking. They also have built-in tools like loyalty and marketing so your best customers keep coming back. And right now, you can get up to $200 off Square hardware when you sign up at square.com slash go slash how I invest. That's SQU-A-R-E.com slash go slash how I invest. With Square, you get all the tools to run your business with none of the contracts or complexity. Run your business smarter to Square. Get started today. Create in the early about CI, the presentation, it will cut it in the right kind of sizes for the slides and so on and so forth. And by click off a button, it's like 98% there. I could give you so many more examples,

but it really, it's part of every single workflow from Monday morning until Sunday night. What was the hardest part about integrating AI into a venture capital firm? Cultural change. Initially, if you would have asked me, I had assumed it was data availability and data quality. So lots of this data is qualitative by nature if you invest in pre-seed or seed. So lots of this data is not available. We learned over time that you can quantify this data. So for example, a university degree might be very qualitative and subjective.

But if you look at it, you can actually rank different universities. And then more importantly, you can also rank different degrees. For example, in Germany there's University of Bonn. If you would just rank them, they are tier two. But if you study math at University of Bonn, it's suddenly Tier 1. The same is true. Say 10 years back, you worked at Google. You would assume this is a Tier 1.

But if you did, say, front office at Google, not sure if this is still Tier 1 as a founder. But if you were like Level 5 engineer at Google for four, five, six years, This is very, very strong. So you could see how you can suddenly classify some of the qualitative data. And this is what I assumed was a problem earlier. Then contradicting data, that was for along my assumption. So if you have two sources stating different information, how can you actually balance the data

and what is the most credible one? So this is called entity matching and deduplication so that you can create a single source of truth. For long, I thought that's the problem. But now if I zoom out, I think all of these technical problems are solvable. If you bring the right people, if you have the right toolset, they are solvable. Something I completely underestimated is the change management and the ability to adapt workflows and make the data actionable. So specifically, to us as a partner, it was very clear that this is the future of venture capital. But if you have an investment professional who had been with the firm for five years and they have,

had one specific way of adding a deal to the CRM system, doing follow-up, adding specific nodes. It was very difficult to tell them, look, there is this new tool. You can just transcribe the coil with granola. It will push with the NCP into our system. You can just use different skills to build the investment memo. Lots of investment professionals actually struggled, and some even avoided it, like proactively said, look, I don't know if this is so useful. I think I'm faster the way I used to

instead of just spending now the weekends learning how to do all of this stuff and what I found is that lots of people are just afraid if they see a console, this is for the developers, this is not for me. I will use my software interface and do simple stuff as I used to do

and I completely underestimated this. So thankfully we are now one step further but I think that will be the biggest blogger for the majority of the investment firms. I think a lot about this. I have a second master's in psychology. I'm a big fan of B.F. Skinner, who populized this concept of operant conditioning, which is basically human beings react to rewards and punishment, almost like Pavlov's dog, but for humans. And I had Ryan Hoover from Weekend Fund and founder of Product

Hunt on the podcast, and he's really changed my thinking because he doesn't look at things as processes or as one-time things. He looks as things as products. So he creates products, and processes in order to streamline a lot of things that he does over time. And it was really influenced on that. But then when I started trying to apply to my life, there's a friction. In other words, in order to automate something, it might take two to four weeks of managing an engineer. So the easy thing is to go and do it.

So go on LinkedIn and do this manual process. I get positively conditioned for it right away. I get a positive effect. The hard thing is to lengthen your reward site. to four weeks and having to go through all this issues. And then after two weeks, you're almost like back to break even. So six weeks in, you're at break even, then the rest of your life you could use these new processes,

these new technologies. But it is hard to get out of this mindset of just doing something today and just getting through the end of the day versus building in a tech-centric way. I couldn't agree more. And I think it's very important to start with the quick wins. So I'll start with a simple project where within 30, 60 minutes, you get the first win.

For example, our founders, they are in the early 60s. And I remember after Christmas when I retested Claudecote Code and the new model came out that completely changed the game in early December last year. I told him, now is the time. You should just use it. You can describe whatever. And I remember a few weeks later, I sat in a board meeting

and like a virtual one. And my partner came in, he knocked on the door and he was like, Andre, Andre, I had a lot. this eureka moment. It's insane. I tested it to do X, Y, and Z. I've never been so productive. Like, everything will change now. And I think that's really the mindset to start with something very simple, to get these quick wins, reduce the time to value, because then you can start to build more complex projects and you can layer it on top. And that's exactly what you just mentioned, because it might take longer to build something to solve the task once. Like,

Sometimes you are just like, I could just quickly do it. Of course. But if you do this task 20 times a week, then after two, three weeks, you're already brave even. And I think that's right, mental model. What you also said from Ryan, and I think that was also an insight that we had, I would call it you need to measure what matters. And if you don't measure it, you cannot improve it. And to give you an example of what that means, we have built a sourcing engine. So first of all, we looked at the status.

We believed we had comprehensive coverage all across Europe back then, as most other investment firms believed. Now we built a system. First of all, we defined what is comprehensive coverage. So we said, we want to see all opportunities where our competitors end up investing. So we defined a list of 200 investment funds, and then we looked at the legal entities and scraped all of the public registers across Europe. So suddenly we could see the full universe of opportunities where our competitors were investing. in and then we looked at our internal system and said, which companies had we been in touch

with? Where did we upload some documents? So we called them the hits. And then we divided the companies that we have seen by the number of companies our competitors had invested in and you get something that we call a hit rate. So we could see that all across Europe we had seen about 70, 72 percent back then. This was 2019, 2020. And we believed we would see everything, but we missed one in four opportunities.

So we introduced the system and we built our platform Eagle Eye to a point where we have, depending on the quarter of the season and so on and so forth, about 95, 96% of hit rate today. Meaning, we see 19 out of 20 opportunities and we introduce a time factor before the closing. So we see 19 out of 20 opportunities across Europe, at least six weeks before the close the funding round. And now we asked our investment professionals, why are you not using the platform? And they said, look, I just meet the other investment professionals, my peers at other firms, and I'm sure I will see them.

So we said at some point, look, we need to introduce the right measures. And we agreed as the partnership that we measure our investment professionals who are split geographically on the more junior levels by hit rate. So suddenly you cover France and you have the hit rate. where you measure your coverage. And I don't care if you use our platform Eagle Eye or not, but just as a hint, Eagle Eye has 96% of the relevant opportunities in this geography, so it might be easier to just use the platform versus going for another coffee meeting with this other VC.

And this is the way how we got our investment professionals to actually start using the Eagle Eye platform. And we made it part of their performance reviews. So we said, look, we don't care how you get there. We look at the outcome, but we will measure your performance. performance among other dimensions by the geographic hit rate. And we have been doing the same across other dimensions and where we see, once we measure it, we know the status quo and we know if we are on track for something. And we can actually measure the impact of any initiative that we have, be that on the tech side,

also the human more traditional side of things. Going back to operate conditioning, you had VCs in your organization, analysts, associates that had been trained, that have been just conditioned to meet with other VCs, and you weren't pushing a process to them. You were pushing an outcome and output, and then you showed them essentially a new skill, Eagle Eye. And over time, as they start to outperform and get positively conditioned for this new process,

now they started to just naturally adapt to this new process. Exactly. Perfectly summarized. Like in the beginning, I thought, I had more of a product mindset. I said if we built Eagle Eye into such a great platform, they will just intrinsically use it by themselves. And I said, look, we see 95, 96% of the opportunities are in the system. So obviously they will use it because they will see that they have a competitive advantage.

They did not use it. So we needed to create the right measures and incentives for them and then show them the tool available to get there. Because otherwise, if we look at France, back then we had whatever, 65% coverage. So we missed literally one in three opportunities in France. The system had 19 out of 20 in there. why not use the system? So you can change the behavior if you set the right metrics and also have the right incentives to achieve these metrics.

Anthony Pompliano on the podcast a couple weeks ago, he's one of these guys that could explain complex things very simply. And he talked about you could turn a loser into a winner, but you can't turn a winner into a loser. And I've been thinking a lot about that. And there's a paradigm in terms of there's a victim mindset, and there's an agency mindset. That's certainly a thing.

But there's also a skill gap. And sometimes, quote, unquote, losers just don't have the skills needed to be a winner. Sometimes they're just using the wrong processes, the wrong technologies. And if you could just give them the right tool set, then they could become winners. 100%. One of the most critical skills across industries is really to unlearn old behaviors. And I think what brought us here might not bring us there. And at times where the technology changes so fast,

like you cannot imagine how often I was between should I use codex or clothe or should I use different kind of allelms. I played around back then with open claw. I hosted stuff locally. So I retested and there is news every week. And I think it's very important if you're deep enough. If you're a beginner, don't let yourself be confused by all of the needs. news out there. Just start somewhere. Like no matter if it's cloud code or codex, they're mostly

the same. They have some differences, but just get started. That's the most important thing, as we said before, try to reduce the time to value and get some quick wins. But if you're deeper into the trenches, I think it's very, very important to unlearn what you learned before if it is necessary and something better is available. And I think just by the nature of how most investment professionals are trained, most of them have a traditional business better. background, they have a banking background, they have traditional MBAs and so on and so forth by education. And I think most have not learned to unlearn some behaviors and relearn new workflows and skills. And that makes them very inefficient. And that also creates blind spots that

oftentimes they are not even aware of. Keeping a beginner's mind as you get older becomes harder and harder every single year. Being bad at something too when I started the podcast three years ago, just doing it publicly and just seeing yourself and just being so disappointed in everything and just wanting to kind of start as a beginner was just so mentally hard

and I just had to grind through it. But the older you get, the more your ego gets bigger, the more you ossify in terms of like new learnings. It just becomes harder and harder. That's why these 22-year-olds are adopting to AI so quickly. Yeah, it's a different form of innovators.

dilemma. Once you become too big, too established and you feel you know it all, it's time for disruptive change. And there are oftentimes the new kids on the blog. So you can look at the YC classes where you have 20-something year old starting companies

and reinventing complete industries. And I think more often than not, they are better positioned because they do not have the existing kind of, it doesn't work. We tried this in the past. it just doesn't work kind of mindset.

You mentioned win rate, clearly an important metric to track as a venture capitalist. What other metrics are you tracking via your AI-driven platform? Look, we track everything, but eventually it comes down. Do we return the money that we want to our... And what's upstream of that? So you can really do the mass. So early stage venture capital funds, typically 10-year fund curation, they can be extended by one or two years.

but let's take the long-term horizon. It is DPI, distribution to paid in. We want to return money, not just paper money. One step earlier, you can look at TPPI, so the total value to pay them. This is paper money, essentially. Now you can look at, and that's an aggregate on the fund level, but you can also look at individual assets.

If we look at the individual assets, we can look at the last valuation, we can look at fair money valuation, whatever we want to apply as a fund. And we can measure this against our initial investment. So we have a multiple on investment. So we can look at this metric. Now we can go shorter and shorter and shorter.

And then we can look, for example, at follow on funding rounds. We can look at the time between follow on funding round. We can look at the step up and valuation. Do they just double evaluation or do they quadruple the valuation? So we can look at all of this. And if we look at an investment basis, we can also look at it from a cohort basis. So we can look at which partner did the investment. How was the investment sourced?

Was it sourced through events? Was it sourced through our network? Was it sourced through the Eagle Eye platform? And then we can try to find patterns. So we might see that one partner is an amazing investor. As often, there are individuals who just have a very strong gut feeling. So they have a good mix of biases that lead to. them to make very, very good decisions.

And we need to understand this as early as possible, try to codify it and really translate that into the broader investment organization. We can look, for example, historically, we had attended more events. So what we did is we attributed the source of an investment opportunity and tried to measure on a cohort basis if these companies are successful. And what we found, at least for us, that we did source very, few opportunities at some of the larger investment conferences. And those investments did not perform as well. So at some point, we just decided, look, we should attend less conferences and do more of,

for example, vertical conferences. We are looking into some of the AI research conferences, ICML, I clear and so on. And we are now looking into other companies that we source at these conferences more successful. So there are lots of metrics that we are tracking, but we need to define in the end what are the key metrics and how do they relate to the input? Because otherwise, who cares if you like one specific investment has an amazing performance, nice, good for the IPs, good for the person who did the initial investment. But if you do not extract learnings from that, if you don't do post-mortems,

If you do not find patterns in this, it will be very difficult to institutionalize your knowledge. And I think that's a big different. Lots of teams build an investment fund. It's a partnership, whatever, handful of partners, few partners, and they raise fund one, fund two, fund three, but they never institutionalize. And some others, and this is what we also try to do is build an investment firm. So we try to institutionalize the knowledge so that it becomes. becomes more independent of individuals. And that's really our approach.

And we leverage technology as one instrument to institutionalize this and make early birth less dependent of humans. Of course, in the end, we know it always comes down still to decision making, deal winning ability, and so on and so forth. So the human factor still matters today. And I strongly believe it will still matter in the future. But the more we institutionalize, the better we can scale it up and give the best kind of characteristics and insights to the more junior investment professionals.

What's compounded the most, as you guys have grown to today, two and a half billion dollars under management? Look, if I think about what got easier over time and what started compounding, it's definitely the brand. So I think that was me being lucky. One of the right decisions I made was joining an established brand because the market has become very competitive and there are new funds evolving, new different models, micro, solo GPs, and so on and so forth, every single day. And I think joining an established platform allowed me to, first of all, see the relevant opportunities. So back then we already had about 10,000 investment opportunities per year that came inbound through our website. So building on top of a very strong network

and inbound deal flow was very, very important. Secondly, it's not only about seeing the relevant opportunities. It's not only about making the right due diligence, but then also making the right decisions. So thankfully, I worked with very experienced partners who back then already did it for more than two decades. So I could really work very closely with them and understand the decision making part. So the decision making is the second part. And now the last part that is also very important is really the excess part. Nobody cares if you saw an investment opportunity, decided to invest, but the first is, founder did not let you in. So you need to be in a position to really win the most competitive

deals. And I think the firm brand is what opens the door. You as an individual partner need to walk through the door. So you need to convince the founder that you are the best partner from early bird in this case. And we all know, like as much as the investment firms try to sell if you work with Fund A, you work with the Champions League out there. And it does not matter which partner is doing the deal, but in reality, we all know it does. Lots of partners are employees. They are hired and they are jumping around firms. So every two years, they have a new firm. But early stage investing is such a long-term business. So for us, it was also very important to incentivize the people long-term so that you also can build these long-lasting relationships,

which I think is very important. So to your question, what does compound? I think the brand as the firm compounds, but also, and we mentioned that earlier on the founder level, we had founder branding. I think the personal brand of the investor also compounds over time. If I want to get in a very competitive AI deal here in Europe, it has become way easier for me after having invested in Alef Alpha, Black Forest Lab, Deepcoats, Nick, and so on and so forth, today versus just five years ago. Do you think the concept of the partner versus the firm is more appreciated by older or more serial founders? Or do you think that first-time founders now because of podcasts and media are starting to

really appreciate that as well? I see a very clear picture that first-time founders tend to optimize a lot for the firm brand. It might be for many of them an ego thing. They might assume that it is kind of a good proxy. So they are associated with success. So if I work with them, it's like the holy gray getting their money. it will help us to be better at talent attraction and so on and so forth. So I see first-time founders optimizing a lot for that.

Second or third, fourth-time founders, I see optimizing a lot for the individual person. So of course, it needs to be a sufficient brand. They would not work with a tier three brand if they have an offer from a tier one brand. But if it is tier one brand against tier two a brand, I have seen many founders who optimize in this example for the better personal fit if the Tier 2 brand had a better personal relationship, the more experienced and the more valuable partner on the individual level. So if I think about it, I think about it as a distribution. So you think about it as like a curve, goes up and down. It has local maximas. It has local minima, but it always goes up and down. And I see every fund sitting somewhere different. So you can see

many funds sitting next to each other. And then you have the maximum on the maximum, you have some of these top brands out there. So, the sequoias, and reasons, light speeds, and so on on the world. And then you might have some other funds sitting down here or up here somewhere in between. Now the interesting part becomes they all have like an amplitude around the curve. And this is where the fund is. And you can have a bad partner that really makes a different on the negative. So it becomes locally lower in quality and it can have a great partner. Now, a bad partner at a tier one firm might lose a deal against a great partner at a tier two firm. And I've seen that too often in reality where they are like very sure they would win a competitive deal because they work for fund X, Y and Z.

But they lose against a partner with just giving their all to win this deal in a tier two firm. Going back to your research, we were talking about these factors that are evergreen that continue to persist. And then there's factors that change from generation to generation. With AI founders today, what has changed about what they're looking for from their venture firm? This comes back again where they sit in the stack. For example, if you have a deep tech hard tech company at the bottom of the stack, as mentioned before, they need lots of, capital, they are looking for how big is your fund? Do you have connections to follow on investors? Is this a great signal for later stage investors? So we can attract capital. And I think we can show

that within early bird, we have an early stage, we have growth funds, we have very deep pockets and can easily invest 50, 60, 70, 80 million over time. I think that's very important. Then if we look one level higher, for example, on the AI labs, on the software infrastructure, structure. For them, it's actually very important that we have experience of commercializing research and technology. So for the AI labs, depending on which modality you work, if it's text, for example, coding, or if it is image, video versus different kind of world models that are in the 3D physical space, the go-to-market motions are the most complex part. So the teams themselves, like, they can pull it off. They have the credibility. And with funding, they're

also get to compute abilities models. But the biggest mistakes that I see being made for many of the AI labs is on the go-to-market side. Do you build a playground? Do you do direct sales? Do we do partner sales? Do you go for frontier? Do you go for forward deployed engineers and so on and so forth?

So I think experience, and that's lots of transfer learning, if you have an existing portfolio in this space already and you can tell them this is how company ABC did it, why don't you chat with the founder over here? incredibly important. If you look at the application layer, I think it becomes a lot about specific metrics, like how do we reduce churn, what are the right metrics?

So think about lifetime value, what does KAC versus LTV look like? And then it becomes a lot, a lot at the application layer about distribution. So in the application layer, we have them with founder branding, for example. We give them experience with different kinds of channels. Why don't you speak to this person?

They have been doing amazing on GEO. or search engine optimization or some of the other channels in the past. So I think the application layer is really a lot about distribution. So it depends. And I think that's also the nice part of the job that it's not one size, it's all. And depending on which companies you invest in, they want to see something completely different.

To give you another example, we have some companies in the portfolio that have a strong thesis on sovereignty. So for example, ISA aerospace, They build a lower orbit rocket launcher, which is about independent access to space for Europe and for some of the other countries. Here it's incredibly important to know the right politicians, to be close to policymakers and so on and so forth. So it's a very specific skill that these companies want to see in their investors. Early Bird was really this early adopter in terms of going from a pre-AI world to now fully-induced. integrating into AI with your engineers and your processes.

Today, arguably, we're in a post-AI world. Where does Alpha accrue to the venture capital firm in a post-AI world? This is an amazing question. I just published a new data-driven VC landscape 2026. So I've been running the data-driven VC community for many years, and I've been writing a newsletter on this topic as well, which is really about the topics that I described earlier, and I've been researching that are very close to my heart, like how can investors leverage

AI and automate their processes, what are the patterns and characteristics of successful companies and so on and so forth. And in the landscape report, we actually differentiate between why are investors leveraging technology? And we can see that about one third is leveraging technology for efficiency, but two-thirds are leveraging it to generate alpha, meaning making a difference. And today, where a majority of the public data, is available. So you can buy harmonic, you can buy pitchbook, D room, whatever these providers

are, but majority of this data is available and structured. I think it really becomes all about proprietary data. And proprietary data, I mean transcripts of meetings. I mean investment mammals. I mean decision making. So for example, at Early Bird, we introduced a post-investment committee survey in 2018. Since then, I think we have captured more than 200 investment committees where every participant of the investment committee rates. How do I like the market, the product, investment team. Would I do the investment? Yes, no.

So we track individual decision making and we can see does that collectively lead to the right outcome. We can measure that against the performance of the companies over time and so on and so forth. And I think this is really the data that is the most unique. No firm will ever be able to sell this. This is unique to early bird. Now, the interesting part becomes if you productize all of that, if you make this available

to replicate judgment, for example, in the upper part of the funnel, so you present investment opportunities to your investment professionals that have the highest likelihood of success based on success scoring model, but also represent the taste of early bird. Which opportunities have a higher likelihood of getting to a positive investment committee outcome at early bird. Every firm has a different taste. Every firm likes different kind of opportunities and we try to codify that. And if we present new investment opportunities to our investment professionals and they don't like it, they might reject it. They say, I reject because I don't like the team. I don't like the market, the product, whatever. So they put in specific labels and we can

continuously retrain our scoring algorithm, our recommendation engine so that we understand what is the taste of early bird. And if you map that, the likelihood of success of a company with the taste of early bird, then ideally you can have a fully automated pipeline, just scheduling meetings into my calendar. I wake up in the morning, I look into my calendar, and I will meet the best founders at the intersection of likelihood of success and fit for early bird. And I will get the meeting prep, send 15 minutes ahead. I can look through that, and I can do a sequence of meetings. They are transcribed. It is pushed into the system. and the system will relearn again to refine the scoring and recommendation for Andre.

This is where I think alpha will be truly generated. Michael Gilroy from Marathon Venture Partners talked about this pre-portfolio concept that they have. Every venture capitalist, whether they want to admit it or not, are making bets on what they're likely to invest in and where they're making those bets with their time. if you could get better upstream in terms of what portfolio companies you're focusing on, then your investment decisions downstream will be better, whether or not you could get access as a different thing, but at least you'll know exactly where you want to invest.

100%. And that's something we try to codify and automate as much as possible. I want to leverage technology and specifically AI to free up time from our valuable investors that we can use to do the real value work, really spend time with the founders, spending time human to human, and not so much sitting there and crunching different sheets and so on and so forth. I don't want to look at the input, move numbers around, look for different kind of research and so on and so forth. I can look at the output right away and can leverage the output for my decision making. It's actually where I was going to go, which is in a post-AI world, do the characteristics of top quartal or top-dec's change versus a,

pre-AI world? Some parts of it, yes, but I think the most critical part remains the same. And this is really decision-making. In the end, you need to be an excellent decision-maker. And we are not paid to make 10,000 decisions a year. We just need to make this one right decision. Because we all know, due to the power law, it is about these few outliers. And if we can ensure, as Early Bird has done in the past, we have raised 17 funds. And thankfully, we had one or more outliers in every single fund in the past 30 years. This is something that you need to repeatedly achieve. And I think this all comes down to great decision making.

So I think great decision making is where we should spend our time. We should free up our capacity with redundant work that does not really create value, but really focus on the decision making. I don't want to reject 2,000 opportunities a year because they are not in our geography, they are too late, we don't invest into this industry, and so on and so forth. You cannot imagine how much time we have spent in the past because we also said, look, our brand stands for trust.

So if we have 10,000 founders reaching out a year through our website, we got back to all of them, every single one of them. And the big question is, is this actually where we create. the value or can we automate this part so that we see these are the 200 most promising opportunities where we know that fit all of the hard criteria and suddenly we can go way deeper and way fewer opportunities. In an ideal world if you zoom out we do in our early stage fund a portfolio of about 35 companies and we invest this across three to four years so say this is 10 to 12 investments per year. If we do 10 to 12 investments in an ideal world,

I only want to spend time for the full year with 10 to 12 companies. The problem is we don't know ahead which 10 to 12 companies these are, which is why we need to look upstream in the funnel. And the question is, how far do you go up? And in the past it was we needed to go up to the highest point. And if there's 10,000 opportunities, potentially there are 500 that are from India out of geography. There are some that have raised already 50 million. That was too late for us.

so I needed to scratch them out one by one. And today I can look way deeper in the funnel because majority of this top of funnel screening work can be done by AI. So I can finally spend my time with way fewer opportunities as close as possible to the 10 to 12 that I actually want to invest in per year, but can go way deeper.

In the past, I spent like whatever, I did four or five meetings with the founder and then we came to a decision. Today I can afford to do 10 meetings with the founder or I can have five. way deeper conversations, way better prepared. So we are not stretching the surface, but I had so much time preparing for this meeting

that we can go to the essence that we matter for the decision-making in the end right away. You mentioned venture capital really comes down to these couple decisions per year. You even said one key decision. How do you get better at making that one key decision? I mentioned before we have this asymmetric cost basis. And the false negatives are the expensive ones. Because I said false positive, I say yes, but the company is not successful.

I lose my money once. False negative, I say no, but the company could return 50 times the money. I lost 50 times the upside. So technically, yes, I just want to do this one decision per year. But I actually need to do the 10,000 decisions a year. Now, the question is how reliable is a no? And this is a big question where lots of people did not trust the algorithm to say no. I actually published a paper in 2020 where we benchmarked human investors against the machine learning model on predicting the confidence of saying no.

And actually the machine learning model was as good as the best investor in our sample, I think from 120 investors. So we got the trust to let the machine say no at the top so that we can focus more at the bottom of the funnel where we can say yes. For this one decision where we need to say yes, I think lots of stars need to be aligned. So firstly, it needs to fit our heart criteria. It needs to fit into geography, stage, ticket size. I mentioned before that we are quite disciplined on shareholding. So it also needs to fit in there. Then it needs to fit into a thesis.

So we are probably more on the thesis-driven end than on the reactive end. Actually, at Early Bird, if you say in a deal floor call or investment committee, oh, this is exciting because there is fund ABC looking at it. It's rather negative signal because we want independent thinkers. I don't care what other funds say. It's nice in the sense of what we said at the very beginning. Great founders who can raise will likely keep this capability later on. So it's great to know this founder can create momentum,

but it should not impact our decision in a way of we should invest only because of this reason. So we should think through it from first principles, and then it depends a lot on the opportunity. It is very different for an AI lab than it is for a rocket company or a fintech company in the application they are. To me, it's kind of a paradox of making this great couple decisions per year because I think there's actually two almost opposite ways to think about it. One is I become religious about this idea as quality is downstream of quantity. I went for nine months.

I did five podcasts a week because I just wanted to get bigger, faster. I wanted this feedback cycle watching myself, seeing the metrics and just constantly getting better. So at some level, quantity is upstream of quality. On the other side is, of course, it's about these couple of key decisions. And I think the other side is actually a capacity, keeping this almost empty mind. It's almost like a philosophical or a Buddhist concept. And the way to actually operationalize that is through different practices. For example, I go to sauna and coal plunge.

It's basically bringing your mind to a level where it's not overworked with the constant kind of minutia of the day-to-day. One of the things that I've learned is that doing mind competes with the thinking mind, Doing tasks, doing day-to-day tasks competes with thinking about the business. The way that I figured this out is on the weekend, I would go biking. And I would always come up with great ideas. And I try to isolate all the variables. Maybe it's the biking.

Maybe it's the weather. Maybe it's being outside. And then I realized, no, it's the weekend because I don't have all these doing things. So that wasn't competing with kind of thinking from first principles. So in some ways, you want to do more quantity. In other ways, you want to do the exact opposite, which is less quantity and kind of leave room to think. This is such an important topic. And I think wherever you start in life with a new topic, be that a new hobby, be that a new discipline you want to understand.

In the very beginning, it's all about getting a reps. So there's this 10,000 hour rule you need to spend 10,000 hours to get to like an expert level in something, be that a new language, be that a new coding, new research field whatsoever. I think in the beginning, also decision making, I was very happy that as a junior investment professional, I got the reps. So I have seen thousands of opportunities and rejected them. And I had a more senior investor, a partner sitting on my side. And we talked through these opportunities in our deal flow calls. So I could really see what good looks like.

But once you're at this point where you know what good looks like, you should not distract yourself with too many things because it will just slow you down and you have a higher chance of missing the good opportunity. So once you have all of the reps, I think it becomes about freeing up capacity, getting into the state of a prepared mind so that whenever you see something great, you're ready to shoot. And this prepared mind also a great concept. It's not just having capacity, it's not just having a free mind. It's also your information diet. What are you learning? you talking to in order to have the mental scaffolding to be prepared for these opportunities. I come.

Yeah. And as you said for the weekend, I have the same for my week. I also, back then because I completed my PhD in parallel to my investment work, back then I had Fridays blocked to do some research work. And even until today, I have parts of my Friday still blocked for reading, reading new papers, reading newsletters, just consuming different information. because that's the input that I need towards the end of the week to process over the weekend when I have full freedom where I just spend time with my family. And I can really think through and process where I'm not running from meeting to meeting.

And I found this to be a great setup for the week where I do all of the meeting from Monday to Thursday, Friday only selective meetings. And then Friday I can consume, prepare everything on input that I need for the weekend to process, come up with great idea and then put them down into a street. structured kind of format and really get into the week for executing again. If you could go back to when you were just in the final stages of your PhD and you just entered venture capital, what is one piece of timeless advice you'd give a younger Andre that would have either accelerated your career or helped you avoid cost of mistakes? I have always done things in parallel. So, for example, I started working as an engineer in parallel to my mechatronics engineering

studies when I was younger. I have also worked full-time in parallel to my master's degree. I have worked end-to-end full-time alongside my PhD. So throughout my whole PhD, which was a full-time PhD at Tia Munich, which frankly speaking is also not something that you just get gifted. So you really need to work hard for that. I worked full-time as an investor at Early Bird in parallel. So I had probably two years of overlap where I did a full-time PhD and full-time work at Early Bird. And I think doing that for more than a decade, these two streams in parallel. I always invested more and more and more. And I had years of 100-hour work weeks plus that really consumed everything of me.

And I think later on in life, I got to a point where I just saw, look, if you are just grinding the whole time, you miss the opportunity to zoom out and really see the bigger picture. So thankfully, I also had strong mentors who advised me, look, Andre, slow down a bit. zoom out, think about what really matters, about personal life, about health, about business, because then you will be in a position to make better quality decisions. And I think that's something I would give as an advice to my younger self to zoom out more often and that just working more and harder does not always lead to a better outcome. People ask me, our startup, a marathon or sprint, I always say both. You need to be good at both

skills. Well, Andre, I think I heard about Early Bird back in 2009 when I first got into a startup world. It's one of the most iconic firms in Europe. So it's a pleasure to have you on the podcast. And thanks so much for sharing your story. Thanks a lot, David. Thanks for the great conversation.