Subscribe
Share
Share
Embed
Conversations with the founders, investors, and operators driving Belgium’s startup scene,
Exploring the ideas, stories, and strategies that accelerate the ecosystem flywheel 🚀.
SÉBASTIEN: My name is Sébastien Deletaille. I'm a tech founder, and today I'm leading an enterprise called Rosa.
ROBIN: First of all, give us a general idea of who you are — your background, what you've been up to over the last 20, 25 years of your entrepreneurship journey.
SÉBASTIEN: Twenty-five years ago I was still at university. But since then, I've had the pleasure of realizing myself as a tech founder. Defining yourself that way is a question of identity, and each word matters. Founder, because I like to create an enterprise — find a problem, build a solution around it. Tech, because that's how I deliver on that entrepreneurial nature. I like the idea that a few people can move continents — that with a small team you can build something that scales and impacts millions of people.
I've been lucky to be involved in several ventures. One was founded as Real Impact Analytics, later rebranded as Riaktr. We were active in telecom big data at a time when big data was just flourishing, so we had a very interesting high-growth curve riding that trend. Today I'm in digital healthcare with Rosa — patients use Rosa to find and book medical professionals.
ROBIN: Take us back to the first company you ever started. Did you always have that entrepreneurial drive? Do you have a technical background you could build with yourself, or did you always need to surround yourself with people to build what you had in mind?
SÉBASTIEN: The first company I founded was a small thing at school. That's one of the patterns with entrepreneurship — the earlier you're exposed to launching ventures, the more likely you are to become an entrepreneur, and the more likely you are to become a serial one. That's entirely my story. I grew up in a family of entrepreneurs, started building ventures at school, continued at university. I started my career in consulting because I didn't feel ready to launch directly, but I didn't stay long, less than two years, before going straight into what would become Riaktr.
At the time it was pretty obvious that if you wanted to become an entrepreneur, technology was the place to be. We were part of a whole wave of founders, Fabien Pinckaers, Louis Jonckheere, and others, a new generation of tech. The problems of that era were things like the beginning of apps, the beginning of data. AI wasn't a thing yet. We were all launching SaaS ventures, and it was an exciting time.
To answer your question directly, I'm not an engineer or computer scientist by training. I'm a commercial engineer, but I've always been comfortable with technical and mathematical concepts. My developers jokingly call me the geekiest of the sales guys. I focus on talking to buyers, to CEOs, convincing them to trust a younger structure instead of a mega-corporate. But I'm still very connected to the product and how it works technically. I have co-founders who are incredibly complementary. At Rosa, my co-founders Antoine Pairet, my CTO, and François-Xavier Orban de Xivry, my Head of Product, are engineers, and the trio works incredibly well delivering a vision in top-notch technology.
ROBIN: You call yourself a serial entrepreneur, but you learn these things by doing. When you started out, there wasn't much of an ecosystem in Belgium to speak of. So how did you learn? Who did you learn from? What was your strategy for compressing the time you needed to build a growth venture?
SÉBASTIEN: The hardest thing as a first-time founder is that you have no idea what you're actually launching yourself into. The unknown unknown is the biggest risk. The other trait I've seen in a lot of my peers, and in myself, is overconfidence in our ability to make decisions in spaces where we have no experience. Often it works out. Say out of 100 decisions a day, maybe 80 to 85 percent are fine. But that 5 to 15 percent that aren't, if it's an important one, can kill your company.
So the skill a first-time founder needs to develop is knowing when to say "I should ask for help" instead of trusting that overconfidence, distinguishing a high-risk decision from, say, a broken printer that won't change your company's fate. Too many first-time founders underestimate that, or get stuck in the overconfidence. With hindsight, on some of the decisions that defined Riaktr, just asking for help would have changed the outcome, performance, exit, all of it.
That's why at Rosa, whenever I hit one of those turning-point decisions, I go to my network of mentors, people who have been there, done that. That's the criteria. I'm not asking someone with just good opinions. Going back in time, I was always keen to ask for help. I received great feedback from seasoned founders like Stéphane Kurgan, José Zürstrassen, and Alexandre Mars. I also got so much humble advice from peers like Sebastien de Halleux, Fabien Pinckaers, Louis Jonckheere, Jan Hollez, Stijn Christiaens, Laura Warnier, and so many more.
ROBIN: So you've built a strong network of mentors. What are the concrete lessons you took from the Riaktr journey, that's a well-documented one, into building and launching Rosa?
SÉBASTIEN: There are a million learnings, but I like to boil it down to what separates Seb the first-time founder from Seb the serial founder.
When I first launched Riaktr, I was convinced success was entirely determined by the team. Don't get me wrong, teams matter enormously, but if you have a great team in a bad market, you'll go nowhere. When we launched Riaktr, we never asked ourselves what market we were entering, whether it was a good one. We were on a wave, surfboard in the water, and suddenly, whoa, great wave. We never asked: is this the wave we want to surf? Will it keep growing? What are we assuming to succeed here? Who's our competitor set? Those questions were absent at launch.
When you look at what separates the very successful serial founders from the rest, it's often their choice of market, the ability to say, this is a problem with real potential, and this is a competitor set I can out-compete. That's one of my main learnings from Riaktr. We had all the funding we needed, an amazing team, probably some of the sharpest minds in data in Belgium, but we'd focused on the big-data-apps market, and that market never really emerged. We grew for a while, then stagnated. That was the cost of choosing a bad market.
I often joke about an entrepreneur I met in Dubai, also selling to telecom operators. He didn't have a great team, but he was in the market of adding novelty effects to phone calls, faking that you're at an airport, making your voice sound funny, fart jokes. That entertainment market was worth a hundred million euros, and he'd captured a huge chunk of it. So there I was, in front of someone about to be a thousand times richer than me, with the smartest engineers building social graph algorithms and machine learning. That's when it really hit me: your choice of market determines whether you're in the top league or not.
Another key learning from Riaktr: be incredibly careful who you go into business with in terms of investors. There's a real gap between investors who promise the moon, say they're founder-friendly, say they're there to make you succeed, and then, when things get hard, the mask falls and you see a different nature. My biggest learning on investor selection: in my diligence, I only talked to the portfolio companies that had done incredibly well. With hindsight I wouldn't do that again. I'm not saying talk to the total failures either, talk to the companies that did a decent job, one-and-a-half to two times return, not the crazy successes and not the total flops. Those are the people who'll tell you the real identity of a fund, its real culture. You want to avoid funds that promise the moon and then point to the winners who probably succeeded without much help from them. The people who did okay, not spectacular, will tell you honestly: was there pressure? Did they help? Were they ignored? Did the fund actually deliver on its promise?
ROBIN: You also invest, in Syndicate One, and in other companies. Is market size and growth rate your determining factor, alongside founder quality and drive?
SÉBASTIEN: One of my favorite thinkers in tech is Des Traynor, co-founder of Intercom. He has an incredibly simple framing: are you working on a big and frequent problem? Two axes, is the problem big or small, rare or frequent? He says the biggest companies in the world are all working on big, frequent problems. If you're working on something big but rare, there's space for that. Small but frequent, there's space for that too. But don't work on small and rare, that's the dead zone.
I've brought this framework back to young founders so many times: are you sure you're working on a big, frequent problem, or is it frequent-and-small, or big-and-rare? That's often where I land on: I'm not sure this is a venture I want to invest in.
That's my lens, I start with market. If there's no founder-market fit, I'm less interested too. But I also accept that no one really knows. In 15 years in tech I've seen people with nothing succeed brilliantly, and the most brilliant people on the best problems go nowhere. There's a form of, I won't call it randomness, I won't call it luck, realization in the market that few people control. And that's okay.
ROBIN: Beautiful segue into Rosa. Talk about your new venture, what are you building, who for, and why?
SÉBASTIEN: Rosa has been one of the highlights of my career for the last six years. We believe healthcare needs to change, being a patient today sucks. That's not a knock on the doctors or the infrastructure in Belgium. It's that the mental load of being a patient is so huge that something has to be done.
That's why we founded Rosa. It happened during a period when my father had a serious medical incident. He ended up in the emergency room, and one morning a doctor came to us and said it was time to say goodbye, that he wasn't going to make it. None of us, my brothers, sisters, mother, were ready for that. We all started crying. Four hours later, another doctor comes and says, "Have you heard the good news?" I said, what do you mean, good news, you just told us he was dying. He said, no, he's fine. I asked, were there new results, a new test? He said no, it was already in his file that morning, the other doctor could have told us he was fine.
That was the moment I confronted the first doctor, who told me, "I always tell families he's going to die, if I'm right, you were prepared, and if I'm wrong, you'll be happy." I said, that's not okay, not just emotionally, but in terms of how you treat patients, their empowerment, their access to information, their ability to make sense of their own data. There must be a better way.
From that experience I asked: how do we get into every patient's pocket? What does it take for patients to trust me with their medical data, so that one day managing your health means Rosa? Framed that way, it's a huge, frequent problem. Becoming the reference patient app for a country like Belgium is a big market, but it's not immediately reachable.
Plenty of people have launched a patient app with a beautiful interface, my data, my appointments, my care team, but click on any box and it's empty, because they don't have enough health professionals, don't have access to the data, or hit some technical wall. Launching a patient app is a hard problem, but it's not really a tech problem, it's a traction problem. Patients don't wake up and search "what's the best patient app in Belgium." They go to the hospital, and the doctor or hospital tells them which app to use. So instead of aiming at the patient, Rosa aimed at the health professional, the hospital. We asked: what's a frequent problem we can solve for them that gets us, over time, to being a patient app? The answer was online booking, a small problem, but incredibly frequent. I love frequent problems: they give you constant feedback and constant chances to improve. If a patient has a bad search once, because they're searching often, you'll get another shot to show them you've improved.
So Rosa's story over six years has been continuous improvement, convincing an initial base of users, plenty of mistakes, but month-on-month growth, non-stop, for six years. That's the benchmark of the world's best marketplaces: demand of patients looking for an offer, supply of medical appointments, matched. We're honored to have delivered that growth and earned the trust of patients and health professionals. Today we're the biggest medical platform in Belgium, we manage a bit more than a million medical appointments a month, and we're still growing month on month.
ROBIN: What's next for Rosa?
SÉBASTIEN: Three things. First, we're in a real regulatory battle to move from being a booking platform to being a full patient app. There's public IT infrastructure in Belgium that controls access to federal services, my vaccines, my blood tests, my imaging results. We have critical mass, user consent, we're clean on security and privacy, we're not monetizing data, we're doing this on the patient's behalf. But it turns out public IT doesn't want other parties, even nonprofits like Rosa, serving patients on this. That's one of our biggest fights right now, because I want to do this by the book, maintaining trust in the system matters too much not to. And I dream of a patient app that doesn't just dump a list of PDFs you have to click through one by one, I want it to show the evolution of your vitals over time, to be solution-oriented instead of data-oriented.
Second: I've realized healthcare IT in Belgium is a world of monopoly, and I hate monopolies. I want to show that a nonprofit approach to tech can fight them, that cooperatives, where users are shareholders, can deliver better outcomes than venture capital. I want to show we don't have to live in a world where a tech monopoly raises prices 15 percent and users have no alternative. That's not theoretical, one of our biggest competitors just did exactly that, and said thank you very much. That's the ideology I want to fight.
Third, eventually, internationalization, though that's the avenue I'm exploring least right now, because there's still so much to do in Belgium. Healthcare IT spend here is more than a hundred million euros a year, just in this small market. Plenty of room to grow before we look elsewhere.
ROBIN: You started ventures before the AI era, and you're building one now, deep into it. How has that changed the way you build the company and the product, and what you see around you?
SÉBASTIEN: It's an incredibly uncertain period, and I tend to ignore people who claim they've figured it out. There's disruption, and it's changing consumer habits, enterprise habits, and founder habits. There's no debate that this changes how you behave as a founder, founders who ignore the shift and keep building the old way might not realize it yet, but they'll get into trouble.
At Rosa we've fully embraced it. All our developers have unlimited access to every AI assistant they want. Our best developers already don't code by hand anymore. That's let us ship faster, but it's also exposed bottlenecks elsewhere in the value chain, testing, CI/CD, other parts of the pipeline aren't necessarily built for unlimited code production.
Two things stand out for me. One: this technology is an invitation to adopt fast and stay fit, your competitors will outpace you if you're not. Two: whenever there's a disruptive technology, there are winners and losers, so see it as a world of opportunity, but don't forget not every problem is a technology problem. That's a fallacy I see in a lot of founders: "I don't need developers anymore, I can ship directly with Lovable, with Cursor." But that's viewing the world purely through a tech lens. What about traction? Brand? Distribution? You can have a great app that nobody ever finds. Every day someone tells me they've built a Rosa lookalike by prompting an AI coding tool for a week, and sure, that's because they think Rosa's problem is a tech problem. It's not. You need to rank on SEO. No AI tool is going to code you that. You need people to think of your brand when they need, say, a pediatrician with a specific subspecialty. No AI tool codes that either.
So all the other things that explain a tech venture's success get ignored because there's a revolution happening in how code gets produced. That's my one warning. It's great that founders can prototype faster, get earlier feedback, test traction sooner, but don't ignore everything else it takes to build a real venture at scale. At Rosa today, most of my problems aren't about adding features, they're about resisting hackers. Every day people are trying to break into my platform, get at my medical data. Or scaling, being available 99.9 percent of the time, handling millions of appointments and reminders. Those are problems you don't see when you're launching. Speed to launch has accelerated massively, but platform health and the rest of it, those aren't easily fixed by the same acceleration.
ROBIN: Let's talk about the ecosystem here in Belgium and Brussels. You've been visible and active, on the board of Startups.be, investing in Syndicate One. How have you seen the ecosystem evolve since your first ventures?
SÉBASTIEN: Similar to other "mafias," the NetLog Mafia, Showpad, Riaktr was an early venture that attracted people who wanted to become tech founders over time. Out of the Riaktr network, more than 50 ventures have launched, including unicorns and companies founded abroad, in the UK, the US, Switzerland. We attracted more than 400 people through Riaktr, and more than 50 of them went on to launch their own ventures, some very successfully.
Building ecosystems is something I care about. At Riaktr our message to people we recruited was: contribute to this venture, and when you're ready to launch your own, you'll have learned from us and your peers, and you'll be able to apply it to whatever problem and market excites you. We were very successful at that.
The hard part is that for these ecosystems to go beyond one founding company, you need more, a funding ecosystem, and talent seasoned in the later stages: not just getting to 10 million, but 10 to 50, 50 to 200. That's where the French-speaking side of the ecosystem struggles most, a lack of people who've actually done the later stages of high growth. Ghent's ecosystem, by contrast, is rich in those profiles, which is part of why so much capital flows there, the soil is richer, richer in experience and in the ability to succeed. I do see some divergence, Brussels losing pace, producing fewer high-potential ventures at the rate it used to. Fewer ventures launched overall. I don't know if it's a post-COVID hangover or something structural. It's an open question for me.
ROBIN: Do initiatives like Syndicate One, network-driven, very Belgian rather than Ghent- or Antwerp- or Liège-specific, help close that gap?
SÉBASTIEN: Yes. When I say ecosystems form at a city level, it's because the city is the right unit of aggregation. Brussels might even be too big for that, crossing from one side to the other takes an hour, so people on the south side aren't naturally drawn to hang out in the center or the east. A city is a good unit, it's easy to meet up, and proximity creates the possibility of serendipity, which doesn't happen when you're an hour apart. Within cities, you need communities that actually create ties, connection, exchange, learning. Ghent has done that exceptionally well, and other cities should take note. But it takes more than leaders and spaces, it takes participants willing to show up, ask questions, share. That's an untold part of the story: not enough people bringing that effort. That's where Syndicate One plays an important role, it has the pull to bring people together, because it brings exceptional members and exceptional speakers, giving people a real reason to show up. I think Syndicate One will end up being one of the reasons Belgian tech accelerated, in both success and in exchange of learning. The open question is whether participants can do more to make these ecosystems truly viral, the flywheel effect.
ROBIN: Final question. You're seeing fewer startups being formed and founded, even though the barrier to entry has never been lower. Is it a talent problem? Is this the same conversation we've been having for 20 years? What needs to change?
SÉBASTIEN: Best possible question, and honestly, I'm clueless. I'm just observing a decline, and I don't know if it's a COVID hangover, a generational shift, or a cycle, because in the history of humanity, in nature, in economics, there are cycles, highs and lows. Maybe we're in a low point. Between the first generation of Belgian tech founders, Bart Becks, Jürgen Ingels, José Zürstrassen, and the second generation I mentioned earlier, there was a dip, two or three years where nothing much happened, and then people re-emerged. Maybe we're in that dip again.
What I see in Ghent is that they're just better at pulling in people who were on the edge and needed a small push to launch, because the ecosystem is strong, the soil is rich, and there have been exits that let people reinvest. Maybe that's happening less in the rest of Belgium. How many people in Brussels have actually had an exit? You can count them. Will there be reinvestment from those exits to plant the next generation of seeds in Brussels? That worries me, because we can't afford, as Brussels or as Belgium, to have fewer people launching. This is an area of the economy where we need as many horses at the starting line as possible. We can't afford people saying "I don't know" or "I'll take a year off traveling and figure it out later." We need more horses, or we become subordinate to other tech powers, American or Asian.
ROBIN: My hope is that this conversation pushes people who are on the edge to actually start. Thank you so much for your time, Seb.
SÉBASTIEN: With pleasure. The only message I want to leave for people who are hesitating: never hesitate to reach out or attend a Syndicate One event. These are moments to access people who've already had the chance to grow successful companies, and who are always glad to pay it forward.