From developer to multi-hat CTO: a conversation with Alexis Delaporte, CTO and co-founder of Wegrow.

Alexis Delaporte has been coding for more than 30 years. At 42, he wears many hats: CTO and co-founder of Wegrow, interim CTO, instructor at an engineering school. His path tells a story of transformation that many tech leads live without always knowing how to name it: moving from the one writing the code to the one setting the direction. With, in the background, a strong belief in full remote, team versatility, and the developer's role in the face of AI.

"I built the website for my dad's company at 16. It was my first freelance contract. I made 50 francs."

The start is classic: engineering school, hired into a consultancy, five years at BNP. Alexis climbs the ladder all the way to project manager. And that's where things get stuck.

"I was in a mid-management position where I didn't have real impact as a manager. At the same time, I had completely lost the coding side that was my passion. I wanted to go back, become a dev again. And I was told: that's not possible."

That sentence triggered an irreversible shift: he goes independent.

In 2011, he lives his first entrepreneurial adventure with a high school friend. A social network around sustainable development. The idea is great, the ambition genuine. The business model, on the other hand, doesn't hold up.

"We never managed to sell it. We must have made a few thousand euros over several years. But that experience got me into the startup world, and naturally that's where I continued as a freelancer."

That's also when he makes himself a promise: never go back to being an employee. He smiles when he says it, because what comes next will prove the opposite.

"What sold me on Arnaud was that side: I don't know anything about tech, I trust you 100%."

The meeting with Arnaud, Wegrow's future CEO, happens on Coder.com. The Tinder for entrepreneurs, as he puts it.

"He pitched his vision to me. At the start, it was really something the two of us did on the side. He still had his salaried job, I had other freelance gigs. I was Wegrow's developer zero."

What sells him is the immediate trust. Arnaud handles sales, Alexis handles tech. Each in their lane, no interference.

"It might sound odd for a partnership, but that's how it worked from day one. And that's how it still works today."

This clean separation even shows up in the physical setup: the tech team is full remote in France, the commercial team is hybrid in Barcelona. Two cultures coexisting, the reflection of a starting intuition that has never been questioned.

Wegrow is a software platform for sharing and reusing best practices for large enterprises. The principle: structuring what worked well into a "recipe" format — here's what I did, how I did it, the results, and how you can replicate it — so that other teams in other countries can do the same.

"SMEs share that naturally. Large groups are spread across the four corners of the world. They systematically do the same thing and can't capitalize on it."

"The beautiful architectures come later."

When you ask him how he built the product, Alexis doesn't talk about an ideal stack or elegant patterns. He talks about speed.

"From day one, I didn't put a beautiful architecture in place. I put something fast in place, something that can be broken overnight and rebuilt differently. Because we know that the product we initially have in mind is never the one that ends up working."

The stack is Java and Angular, fairly classic. The strong technical choice was deliberately tight coupling between the data model and the presentation. In theory, that prevents the system from evolving well. In practice, it lets you wipe out an entire part of the application and rebuild it in a day.

"For the first three years, we could deploy to prod three or four times a day. We were developing very fast, deploying very fast, with an extremely fast feedback loop."

And then, inevitably, the moment comes when speed isn't enough. Regressions pile up. Stability becomes more valuable than raw agility.

"We won't detect it ourselves — the structure will tell us. The moment we start to have more pain than benefit, that's when we need to refactor. Piece by piece, obviously. The Big Bang rewrite very rarely works."

"At one point, I decided to hire a CPO so I could focus on tech."

That's the structuring decision in Alexis's journey at Wegrow. The one that changes the team's dynamics.

"I no longer wanted to think about product at the same time as thinking about tech. I wanted to focus on architecture. The CPO has full authority over the product and the roadmap. We work hand in hand, but each handles their scope and their team."

Today, the tech and product team is about fifteen people. On the tech side: almost only developers. On the product side: data analysts, QA, UX, product management. An organization built for complementarity.

AI in the product: structuring the invisible, questioning what exists

Two funding rounds — €1.5 million and then €7 million — have allowed Wegrow to transform. The second one, in particular, opened the door to artificial intelligence.

Alexis is transparent: he wasn't a pioneer on the topic.

"I'll admit it took me a while to truly grasp what AI could do. At one point, the investors and our CEO told me: Alexis, you're going to need to dig into this. It's a classic. Someone walks up and says: integrate AI. What do I do, how do I do it? No idea — you integrate AI, that's it."

The breakthrough comes through hiring. Alexis hires a staff engineer who spent two years at Total's R&D lab in the United States doing AI with Google. It's with him that the first building blocks fall into place.

The first block: turning unstructured data into structured data. Best practices already exist at the clients' offices, but they're buried in PowerPoints, Word files, SharePoints. AI detects and structures these contents.

The second block, at the other end of the chain: letting users ask Wegrow a question in natural language and get an answer based solely on the existing best practices. No hallucination, no invention: the answer is grounded in the structured data already in the database.

"Our big advantage is that our data is structured. Today, LLMs rely on SharePoints, intranets, unstructured data. The answer can't be as high quality."

"A developer's job is to turn a complex problem into simple logic."

AI for coding — Alexis and his team have fully adopted it. First GitHub Copilot, in assistant mode. Then agents that write the code on their behalf.

"It's a real game changer. It lets us go much faster, and it's relatively faithful to what we want to build, as long as it's well guided."

But not all developers took the leap at the same pace.

"Seniors are the most resistant. Juniors arrive saying: this is amazing, we can do anything. The seniors temper it: careful, you're losing control. And that's the value of having both. At some point, they convince each other. The AI cowboys settle down, and the resisters end up using an agent too."

On the question stirring up the entire market — will the developer disappear? — his answer is crisp.

"The one who just writes code is in for a rough time. They're going to be replaced. But a developer's job isn't to code. It's to turn a complex problem into simple logic. And that will always exist."

He pushes the reasoning further:

"Today, everyone thinks they can build their own little software without developers. And it's true. It works very well. But large enterprises don't work like that. They went from SME to large enterprise thanks to industrialization and standardization. Having 15,000 little pieces of software each running on their own machine isn't standardization. The day they want to standardize all of that, they'll call a developer."

"The first one we couldn't replace was me."

Alexis's most personal transformation is that of the CTO who lets go of the code.

"Going from CTO co-founder who's the first developer to a visionary CTO who no longer codes, that transformed my view of the role. I had become too central to the project. Wanting to keep my hands on what I had created was natural. But it's also a bit of ego and pride."

He gets coaching through a program — The Unicorn CTO — which helps him reposition himself.

"I tried to put my vision down on paper. A developer's guide laying out how we develop Wegrow, how you grow as a developer with us. In full remote, there's also that side of career management being more complicated. There's no natural visibility when you're hidden behind a screen."

His philosophy on growth rests on versatility rather than pure specialization.

"The staff engineer isn't a super-developer better than the others. They're someone with a broader range of skills, and who, with all those skills stacked together, handles more complex problems because they have a more global view. Expertise lets you go fast. Versatility lets you be resilient."

The ultimate proof: four months of paternity leave with nothing collapsing.

"I had documented my vision, given the decision matrices to the team. The team kept running, kept making decisions, and the product kept moving forward during my absence."

Interim CTO: walking in when it's on fire

Today, Alexis puts that ability to make himself dispensable to work for other companies as an interim CTO.

"At Wegrow, things are humming. I've been building something for six years. I needed to be confronted with organizations that hum a lot less well. I show up at a client when it's on fire. I'm there for three to six months, to put the project back on the rails and make sure it can live without me."

The cases vary: a CTO who's a developer who no longer knows how to position themselves in front of a growing team. A startup that has rediscovered its market in Africa but no longer has anyone to run the product. A CTO departure to anticipate during the months of recruiting their successor.

The most stressful case? That startup with no team and a customer knocking at the door.

"There's both the urgency — the product is being sold — and the void. There's no team to provide support and deploy. You have to onboard people very quickly and make them autonomous."

Against expectations, his first move isn't to put people to work.

"The first thing I told them was: you're going to train. You're going to train because there's technical debt to fix. And to make sure it's done well, you're the first lever."

And to win the trust of an existing team?

"I don't get my hands in the code. The CTO isn't the super-coder. I won't talk about code, but I'll talk about architecture. And it's mostly about motivation. If you can motivate them, you've earned their trust."

Remote: fewer pointless meetings, more efficiency

At Wegrow, full remote is 98% of the year. The team gets together in person two days every three or four months.

"The management courses we got were oriented around in-person work. We never had, when we were students, a course on how to organize when you don't see each other. So there's something new to invent every day."

On the classic argument that meetings are less effective in remote, he flips the perspective:

"When people are invited to meetings where they have nothing to do, in remote they turn off their camera and become ghosts. What people criticize as being because of remote, I'd argue is remote highlighting useless meetings that have no reason to exist. In remote, naturally, you have far fewer pointless meetings."

Hiring, on the other hand, requires a clear stance.

"I've stopped hiring Parisians. We've adopted regional pay. And the Parisian who's willing to commute to La Défense for €10K more, let them go. They're not made to work with us."

Hire a mindset, not a stack

On hiring, Alexis is categorical: the stack is no longer a barrier to entry.

"Today, the AI is the one writing the code. The language is no longer a criterion. What makes the difference is the developer's mindset and how they manage to take ownership of a complex topic and turn it into simple logic."

At Wegrow App, the process includes two interviews. The first with him, to detect whether the candidate is reciting or has actually understood what they're saying. The second with developers, in live code.

"We give them a deliberately improvable codebase. We ask them to improve it. And the main rule is: think out loud. Tell us what you're doing and why. Hesitations aren't disqualifying. They show that they're really asking themselves what the best way to code this is."

2026: Ops will take more and more space

His view of the market is clear-eyed.

"2025 wasn't a good year for anyone. I don't think AI replaced jobs. The year was tough and there was less budget to hire. AI replacing people is a pretext to justify not hiring much."

He senses the start of a rebound in 2026, but with caution.

"I don't hire a new person until the previous one is fully integrated. And I think it can take longer now to integrate someone with the new tools."

His conviction for what's next: Ops will become central.

"AI will introduce a very non-deterministic side to development. That side will need to be offset by much more determinism and stability on the Ops side. Monitoring how the software runs is going to become more and more critical."

Conclusion: a CTO who made himself dispensable to multiply his reach

What stands out about Alexis Delaporte is the consistency of his path. Every step — freelancing, the failed first startup, building Wegrow, gradual delegation, the interim CTO work — fits the same logic: understand a problem, solve it, document the solution, and move on.

He teaches because he wants to share his passion. He does interim work because he needs to confront chaos. He stays at Wegrow because what he built keeps growing.

His vision of tomorrow's developer is both reassuring and demanding: code isn't going away, but the one who only codes is. The real skill is the ability to think in systems, simplify complexity, and make others autonomous.

A CTO who no longer codes, who doesn't manage full-time, who gives lectures and puts out fires for others. A CTO who has understood that his best contribution is no longer being indispensable.