In August 2025, I moved into a hacker house in Austin, Texas, with five other founders. I had flown to the US to get a slap of ambition. I got one, just not the kind I expected. At Capital Factory, the biggest coworking space in town, I would introduce myself as the founder of a marketing agency for French law firms, and people looked at me like a man from another era. Everyone was building SaaS. Everyone was building with AI. Nobody was running an agency.
One night, one of my roommates asked me if I knew n8n. I didn’t, really. So I gave myself a week to play with it, with a deliberately silly goal: automating LinkedIn carousels. One week became two. I got hooked, and through that door I discovered building with AI.
Around the same time, a lawyer from New York reached out about his LinkedIn. Working with him, I realized something that four years of agency work had hidden from me: I did not need to master his field to run his communication. If he supplied the raw expertise, the articles he read, the changes he noticed in his practice, I could turn it into something his audience wanted. Expert knowledge had just become communicable at scale. People whose expertise was locked behind billable hours could suddenly reach everyone.
The second discovery
That was the first one. The second discovery came from a much less glamorous place: a CRM.
Back in France, we were building the intake system of a law firm we work with. Intake is one of the unsexy secrets of firms that grow online: what happens to the ten people who contact you today, including the seven who are not ready to sign. Co-building that system meant sitting inside the daily reality of the job. And the daily reality of the job is this: a large part of practicing law is administration wearing a suit. Filling documents. Filing declarations. Chasing information from one system to another.
At one point, we built a working prototype for one specific type of court filing. It took two weeks. It gives the lawyer back one to two full days per week. Read those numbers again: two weeks of building, against one to two days recovered every single week, forever.
The reaction I did not expect
But the moment that convinced me I was looking at the future was not the prototype. It was the lawyer’s reaction. He was not afraid. He was so enthusiastic that he suggested we open a version of it to the public, so that people could start the procedure themselves, without him. A lawyer, pushing to give away part of his own job.
The five links
The story everyone tells is that lawyers fear being replaced by AI. Here was a lawyer working to make part of his own work unnecessary, and he was right to. He had understood the real mechanics, and they hold in five links:
- When advice gets cheap, the market does not shrink. It widens.
- People who would never have called a lawyer, because it was expensive, intimidating, or both, start asking legal questions. The bottom of the market explodes.
- What remains scarce is everything AI does not do: judgment, strategy, standing behind a decision and bearing responsibility for it.
- So lawyers are not being replaced. They are being repriced, upward, on the part of their work that was always the most valuable.
- And if judgment is the product, the scarce skill is no longer drafting. It is being found, and being chosen. Acquisition becomes the bottleneck.
That last link is my old job. I spent four years running acquisition for dozens of French law firms. AI did not make that experience obsolete. It made it the entry ticket.
The firm that does not exist yet
Put the pieces together and you get a firm that does not exist yet in France: AI in production, humans on judgment, systems on acquisition, designed that way from day one. Not a law firm that uses AI. A firm you simply could not have built before. In the US, firms and tools built on this model are raising serious money. In France, to my knowledge, nobody is building it seriously.
That gap is the opportunity, and it is bigger than France. My plan is to use the French market as our laboratory: prove the model here, on one of Europe’s biggest and least digitized legal markets, write down every playbook, then take it wider. The firms are French. The ambition is not.
So that is the bet, and we changed how we work to match it. Hendy still runs its agency, billing services like any agency does. But the real conviction lives in Hendy Studio: a handful of projects, four today, where we take a percentage instead of a fee. We invest our work because we believe in the outcome. That is the difference between billing a client and betting on a firm, and it changes every decision you make.
Why me
I have spent my career explaining new technology to people whose job is not technology. For Google, I criss-crossed France by train for years, training more than 10,000 small-business owners in digital marketing. At Station F in Paris, the biggest startup campus in the world, I supported startups on behalf of Google for Startups. Lawyers are the hardest audience I have ever faced. That is exactly why I chose them.
What you do with this
If you are a lawyer, the question is not whether AI will reshape your firm. It is whether you will be the one holding the tools when it does. If you are a builder or an investor, look at where the value is migrating in legal, and notice how empty the French side of the field is. And if you just want to watch, that works too: I will document the whole journey here, one essay a month, wins and failures included.