Context & challenge
It all started with a simple observation: my kids refused to do their conjugation exercises. Too boring. But they spent hours on Duolingo learning English — drawn in by the rewards, the visible progress, the gamification.
I looked for an equivalent for French conjugation. I couldn't find anything that matched what I had in mind. A few quick interviews with other parents confirmed the problem was shared — and that nobody had found a satisfying solution either.
The opportunity: an open niche at the intersection of gamification, French conjugation, and primary school children. And the perfect excuse to launch a side-business while testing a hypothesis I was genuinely curious about: is it possible to build a complete, production-quality web product by steering Claude Code, without being a developer by trade?
• Role : Solo founder (product, design, development, marketing)
• Duration : Since April 2026
• Scope : Full PWA (front + back + infra) + multichannel marketing
• Stack : React · TypeScript · Vite · Tailwind CSS · Supabase · Vercel · Stripe · Resend · N8N · Framer Motion · Zustand · Phosphor Icons · Claude Code
• Key focus : AI-assisted development · Design system · Gamification · Back-to-school growth
• Results : Live product on bravio.fr · Paying subscribers · Validated beta testers · Back-to-school 2026 launch in preparation
• Results : Live product on bravio.fr · Paying subscribers · Validated beta testers · Back-to-school 2026 launch in preparation
Phase 1 - management & Leadership
Guerrilla interviews with parents to confirm the problem existed beyond my own experience. Positive signal: real need, no satisfying solution identified.
A deliberate learning phase with Claude Code to understand the capabilities, the limits and the right collaboration patterns before building anything.
First functional MVP: French conjugation, 5 tenses, 60 verbs, 4 dynamically generated exercise types (around 3,000 exercises with no manual writing), spaced repetition algorithm. My own kids as first users.
Phase 2 - Design system & rebranding
Two problems emerged simultaneously, making a full rebranding unavoidable.
First, the initial visual identity was built around a time-travel universe — coherent for a product focused on conjugation (past, present, future, imperfect…), but too restrictive for a platform that would integrate mathematics. There's no way to fit multiplication tables into a spaceship.
Second, Claude Code can generate other applications with the same visual style. The MVP had a generic, undifferentiated design that wouldn't survive the slightest competitive comparison.
I built a Bravio-specific design system from scratch, independent of any narrative theme and applicable to all current and future modules — entirely with Claude Code, enforcing strict spec constraints. The new identity was validated by a pool of beta testers (Bordeaux, Paris, Toulouse) before being rolled out across the entire app.
Phase 3 - Product expansion
Based on early feedback: new French and conjugation exercises, a Duolingo-style word bank component to vary interactions, and above all a complete mathematics module (4 operations, CE2 to CM2 levels). The latter required a full architectural refactor to move from a single-subject app to a multi-subject platform.
Progressive reward system: 48 collectible cards unlocked through progress, mastery badges, celebrations with no guilt-inducing streaks. Full Stripe integration with a 14-day free trial (no credit card required) and a hard paywall on expiry.
Phase 4 - Automation & marketing
Bravio being an experimental business site, every recurring task is an opportunity to test and automate rather than do manually.
On the Resend and Supabase side: a 7-email automated onboarding sequence (D+0 to D+10), automatic sync of new signups to the email audience via Edge Function.
On the N8N side: workflows for AI-orchestrated blog article generation and publishing, and AI agents that analyse each child's results to automatically suggest new exercise series adapted to their level, or to re-engage them if they haven't played in a while.
Back-to-school 2026 preparation: rebranding announcement video, Instagram and Facebook content calendar, local guerrilla marketing, printed materials.
What I learned
Working in product has always meant wanting to build, test, discover and learn. That curiosity is what pushed me to train as a full-stack developer years ago — even though design always pulled me more. It's also what made me fall in love with the first no-code tools like Bubble: the ability to create products autonomously, without depending on anyone else.
What Bravio confirms is that the technical barrier has now completely collapsed. You no longer need to choose between designing and building. And while the design barrier is starting to crack too — which can feel unsettling — it's ultimately exciting, because it opens up entirely new spaces for innovation. The silos that used to separate each discipline are falling, and that creates room for a new kind of builder: someone who thinks across the full stack of a product, from the first pixel to the database query.
But the collapse of the technical barrier doesn't just mean you can build. It means you can ship faster, test faster, and iterate, pivot or kill a project with far less friction and far fewer consequences. The cost of being wrong has dropped dramatically — which makes the cost of not trying even harder to justify.
The same shift is happening on the marketing side. Acquiring new skills, testing a content strategy, running a campaign, building an audience — things that used to require entire teams can now be explored solo, quickly, and cheaply. Bravio has been as much a marketing sandbox as a product one.
Visual differentiation still matters — the custom design system is what makes Bravio look like Bravio, not a generic AI-generated app. But the real shift is this: the question is no longer "can I build it?" It's "what's worth building?"
What's next ?
The backlog grows with user feedback, which gets richer every week. New subjects, new exercise types, deeper personalisation, richer parent dashboard — there's no shortage of ideas. They get prioritised based on what users actually report, not what seemed like a good idea upfront.
But beyond the product itself, Bravio will remain what it has always been: an experimentation ground. Every new feature is an opportunity to test a tool, an approach, an automation. That builder's curiosity is what brought the project to life — and what will keep driving it forward.
