From $100M+ raise to liquidation: what Ynsect taught me about scaling deeptech – Jean-Gabriel Levon

In this episode, Antonin Léonard sat down with Jean-Gabriel Levon, co-founder of Ynsect. Ynsect was one of the largest fundraises among industrial start-ups of the 2010s: a decade of ambition, factories and hundreds of millions raised, before the adventure came to an end in late 2025. Jean-Gabriel was one of the four co-founders, the one who "always stayed close to the ground": from the R&D bench to the launch of the Dole pilot plant, all the way to the giant Amiens facility. President, CEO, site director, head of the 4.0 digital factory, then first impact director, he held nearly every operational role in the company. In this conversation, he looks back on the journey with clarity and without bitterness. How a group of friends who started with an insect restaurant ended up raising a million euros when that was almost unheard of, then ten, then over a hundred. Why the jump from Dole (a few hundred tonnes) to Amiens (×50 to ×100) was an industrial leap into the unknown. And above all, the hardest lesson he learned: in industry, pouring in more money doesn't make things go faster. Time is time. Jean-Gabriel left Ynsect in late 2022 to retrain as a carpenter, even working on the reconstruction of Notre-Dame de Paris. He now advises industrial scale-ups as a fractional COO, with one conviction: to carry the voice of operations, the one that says "it works" or "it doesn't", in boardrooms too often dominated by financial profiles. In this conversation, we explore: → Why "time is time, not money" in industry: an insect has a 3-month cycle, so running three experiments takes 9 months, and no fundraise shortens that → The process book, the backbone of a factory: when R&D keeps reshaping it 10 times while you're building Amiens, it creates enormous "noise" → Why industrial hypergrowth is paradoxically slower: complexity grows exponentially with scale, and unwinding technical debt costs about 4× what it cost to buy in → The under-representation of operations in governance: finance and sales "speak the same language", while those who know whether it actually works are structurally outnumbered on the board → The optical illusion of scaling: believing a hardware project can move as fast as a web scale-up, when it's closer to a 15-year pharma scale-up (Tesla, Northvolt, the data centers behind Instagram) → What ten years of the adventure leave behind: ~400 jobs, around thirty patent families, real scientific breakthroughs and an ecosystem that lives on → The forest metaphor: Ynsect as the humus from which new companies are already growing (Kepréa is taking over the Dole pilot), and the "Ynsecteurs" community that still shows up at farewell drinks five years on → What he tells industrial founders today: to de-fantasize hypergrowth and find an investor who is a specialist in their industry ("I'd have loved an entomologist investor") Learn more: asterionventures.com