People sometimes ask me why I keep starting companies. It is a fair question. Building one company is hard enough. Building several, across different sectors and different regions, can look like a lack of focus from the outside.
I want to answer it honestly, and then do something founders rarely do in public: list what I would do differently.
Where it started
I started my entrepreneurial journey at 21. I was trained as an engineer at CentraleSupélec in Paris, and an engineering education teaches you one useful habit: you do not trust an idea until you have built something that works.
I also spent time in more structured environments. Three years at Vinci Énergie showed me how large organisations actually deliver complex projects, with all the coordination, safety constraints and patience that requires. Then three years as a strategy consultant taught me how decisions get made at the top, and how often good strategy dies in the gap between the slide deck and the field.
Those six years were not a detour. They gave me a vocabulary for both worlds: the operator who has to make things work on Monday morning, and the executive who has to decide where to put the money. Most of what I build today sits right between those two.
Why I keep building
The simplest answer is that I believe in a sentence I repeat often: ideas, no matter how brilliant, are mere potentialities. It is the execution that brings them to life.
I have met many people with better ideas than mine. What I have seen far less often is someone willing to spend two years turning one idea into something real. Building companies is my way of staying on the execution side of that line.
There are three more concrete reasons.
First, problems I cannot unsee. Many of my companies started from watching people struggle with something that should be simple. With Maison des Infirmiers, the problem was that home-care nurses in France who want to open their own private practice face a wall of administrative and practical questions, often alone. A turnkey network made sense. Tools like Soinvite, Previzo and RemplaIDEL followed from the same logic: once you understand a profession's daily friction, you see several adjacent problems at once.
Second, technology shifts that open new doors. AI agents are one of those shifts. That is why I founded ThalerTech Group, and why ThalerTech builds custom AI agents and trains teams to use them. I wrote more about what really works with agents in AI agents in business: what actually works in 2026.
Third, the work I am proudest of. Developing a low-cost CBDC solution for a central bank, a first of its kind, taught me that small, focused teams can deliver infrastructure that people assume only giant vendors can build. Once you have seen that, it is hard to stop trying.
What I would do differently
This is the part I find most useful to write down. None of these are clever. All of them cost time when I ignored them.
1. Narrow the first customer even more
Early on, the temptation is to describe your target market generously. "Small businesses." "Healthcare professionals." "Companies in the Middle East." These are not customers, they are continents. In my experience, narrowing the first customer to a single, specific profile makes things move faster: the product got simpler, the sales message got clearer, and feedback became actionable.
2. Ship the boring version first
Engineers love elegant systems. Customers love things that solve their problem today. I would push even harder to ship the rough, slightly embarrassing version early, and let real usage decide what deserves elegance.
3. Hire for ownership, not for the CV
Some of the best people I have worked with had unusual paths. Entrepreneurs come from everywhere, and so do great operators. I would spend less time reading CVs and more time giving candidates a real problem and watching how they take ownership of it.
4. Protect focus inside a group
Running several ventures is only sustainable if each one has a clear leader and a clear scope. In my experience, problems appear when responsibilities overlap, or when the founder stays too close to decisions that someone else should own. A group should share infrastructure and lessons, not blur accountability.
5. Say no faster
Opportunities arrive constantly, especially once people know you build things. I have learned that the cost of a "maybe" is often higher than the cost of a clear "no". A quick, respectful no keeps the team focused and keeps relationships healthy.
Why I will keep going
Beyond the companies I founded, I have backed and advised 12 companies as an investor, across LegalTech, FinTech and HealthTech. Sitting on the other side of the table confirmed something I suspected as a founder: the plan matters less than the person executing it. I explain this in more detail in Invest in founders, not slides.
I have also been involved, in passing, in the strategic creation of political movements in France. Very different world, same lesson: structure, clarity and execution decide whether an idea survives contact with reality.
What I value most is supporting people who start from scratch. Not everyone has a network, capital or a famous school behind them. My approach is rooted in authenticity and humility, partly because I know how much luck and help go into any success, including mine.
I do not build companies because it is comfortable. I build them because I enjoy the moment an idea stops being a conversation and becomes something people use. Whether it is an AI assistant like Nino, a support network for nurses, or a financial infrastructure project, the satisfaction is the same.
If I could give one piece of advice to my younger self at 21, it would be this: you do not need to be the smartest person in the room. You need to be the one who keeps showing up, ships, listens and adjusts.
Key takeaways
- Ideas are potential; execution is what creates value.
- Corporate and consulting experience is useful training for founders, not a detour.
- Narrow your first customer more than feels comfortable.
- Ship the simple version early and let usage guide refinement.
- Hire for ownership, protect focus, and learn to say no quickly.