A physicist by training. 3 decades of building things by hand, from Paris to Beijing to Berlin to Lisbon. Successfully led digital transformations for global companies while building ventures that prioritize human connection over pure tech.
1974 Birth.
1984 Delved into coding.
1999 Failed my first startup in science popularization.
2010 Co-founded an IT services company, Paris/Beijing.
2017 Led a transformation plan for SwitchUp in Berlin.
2024 Launched Nook.coach. Where conversations shape healthier habits.
Jun 2026 Shut it down. I could build it. I couldn't sell it.
Throughout my 3-decade career, I've mastered the art of leading through the intricacies of technology and management, accumulating vast experience in strategic planning and hands-on technical development. Unlike the traditional high-level strategizing typical of CTO roles, my passion lies in immersing myself within the technical trenches, working alongside talented teams to design and execute cutting-edge solutions.
I am eager to apply this blend of strategic leadership and technical expertise, aiming to drive innovation, enhance team capabilities, and lead projects that push the boundaries of what's possible in technology.
This approach is more than just management; it's about staying intimately connected with the pulse of technology and nurturing a creative, hands-on environment that turns ambitious ideas into reality.
I landed in Beijing in 1997, at the university. A year later I joined the French Chamber of Commerce to help French companies find their footing in a market almost none of them understood. I organized training sessions for expatriates, helped budget new projects, and, on the technical side, built the office's first network from scratch. Administration by day, cabling by night; in hindsight, a fair preview of the career that followed.
In 1999 I left to build something of my own: Popsciences, a website covering scientific research, from market studies to product spec to the hunt for financing. It failed. I was 25, I didn't yet know what I didn't know, and the site never found its footing. It was also the first time I understood that building the thing is the easy part.
By 2001 I was back in Paris, deputy project leader for business intelligence at Unilog, a company that by 2005 was posting €593.1 million in revenue with 6,700 people across Europe, before LogicaCMG bought it in 2006.
From 2005 I moved to Micropole, leading BI and e-business projects for Generali, running a team of 12 inside a 75-person department, on projects that at their largest pulled in 100 people. The work was banking and insurance data: BNP Paribas, Crédit Agricole, Generali, Natixis. I like this kind of work. Structured, exacting, high-stakes: it's where I learned to manage scale before I had the title to justify it.
In 2010 I went back to China as CTO and Managing Director Asia for Second Bureau, a company I co-founded, but where the job was running engineering: agile delivery of e-commerce, m-commerce, and social-marketing platforms for local, state-owned, and international clients building their first real digital presence.
I hired and ran a team of 17 engineers. Over 3 years, one person left, against an average 6-month turnover in China at the time. I don't know a cleaner number to point to when someone asks what I actually do as a manager.
In 2013 we shut the company down. It was profitable. That wasn't the issue. The issue was that we'd built two companies that never fully merged into one: strategy out of Paris, operations out of China, with almost no synergy between them. I let the team go, down to a single client, and kept just that account running while I figured out what came next.
What came next, starting in November 2015, was a part-time Hands-On CTO role at Propulsion Ecommerce, a data-driven, cross-border e-commerce operation running out of Beijing, Shanghai, Paris, and Madagascar. I left China for good in 2017.
In 2017 I joined SwitchUp in Berlin as Hands-On CTO, a startup building a smart assistant that manages customers' energy contracts and keeps hunting the market for a better deal. By the time I left, the platform served 200,000 regular customers and could optimize up to 30,000 plans a month. The company had grown past what its Rails monolith could carry, and I spent 6 years leading the shift to a microservices architecture built for high load.
Architecting a future-proof platform, building the multi-year roadmap, hiring the engineering team, and keeping the legacy system alive while all of that happened at once: that was the job, roughly in that order of difficulty. Ruby, Rails 4 through 7, AWS, Heroku, Redis, Tailwind, JWT: 6 years of choosing what to fix now versus what to design for later, on a budget that never quite matched the ambition.
In December 2023, I became Head of My Own Adventures: full-time AI exploration, no deadlines, the occasional existential crisis about AI taking over the world included at no extra charge. It's turned into something like 10 different apps and experiments; more on those soon.
A year into My Own Adventures, that turned into Nook: a health-coaching platform I built essentially end to end, in Lisbon, over 2 years: architecture, backend, the product itself. The hard problem was memory: giving a coaching AI a way to remember a client across sessions without forgetting what mattered or inventing what didn't. I designed a layered system: directive memory for explicit preferences, append-only persistent memory for confirmed facts, vector-based contextual memory with fading and importance ranking for everything else. It was running well before memory became a standard feature on ChatGPT and the other major assistants.
I shut Nook down in June 2026. I'd built something that worked; I just wasn't the right person to sell it. That's a distinction worth being honest about.