Senior Product Engineer

WHY THIS ROLE, NOW Vocca is 2 years old. Our AI answers the phone for medical practices so they never miss a patient call again, and it already handles 70% of calls without a human. Revenue grew 3x in the last 6 months and we are opening the US. The product works. The next 12 months are about scaling it, and this role decides how fast. THE PROBLEM Every medical centre has its own rules. Who can book what, with which practitioner, at what time, what to do with an urgent case at 6 pm. Today these rules live in prompts and in the heads of our ops team. Each new clinic takes hours of manual work, and the ops team is the bottleneck of our growth. We are building the platform that removes that bottleneck. A scenario engine where the behaviour of the agent is structured configuration, not free text. A self-serve onboarding where a clinic connects its software and sets itself up. A dashboard where the medical staff stop executing and start supervising: the agent handles the volume, humans check what matters. This is the difference between selling a service and selling a product. It decides how fast Vocca can grow and how many clinics one ops person can support. The long term vision is simple: a clinic signs, connects its software and is live within days, without a single developer involved. You would be building that. WHAT YOU WILL OWN
• The architecture of the scenario engine and the configuration model. One product that handles the rules of thousands of medical centres without becoming a swamp.
• The self-serve onboarding, from the backend to the dashboard. A clinic deployment already went from days to under an hour thanks to internal tooling. The next step is zero developer ticket on a standard onboarding.
• The dashboard our clinics use all day. Fast, reliable and clean enough that a medical secretary prefers it to the phone. We take frontend quality seriously and we have just created a product designer role to raise the bar.
• The product decisions on your scope. You build the platform roadmap with the product manager, the ops team and the clinics themselves. You decide what we build first, what we build simply and what we do not build, and you prove it worked with a number.
• Reliability in production. When the dashboard or an onboarding breaks, you own the fix and the reason it does not happen again.
• The next hires. You take part in hiring the engineers who join after you, and you help them ship in their first week. HOW WE BUILD
• We are experimental, not R&D. You test, you prove with data, you ship. Challenging a decision means coming back with a number.
• We ship the version that works, then fix the last 1% when it shows up. With AI, this trade off is much cheaper than before. But the judgment about what not to touch is yours.
• We use AI seriously. Agents in every workflow, Claude Max provided. A migration estimated at 3 months was done in 2 weeks.
• Every change lands in front of the medical staff who use the product all day. You talk to them, to the ops team that deploys clinics and to the sales team that sells the product. STACK TypeScript everywhere, backend and frontend (Node, monorepo), PostgreSQL, a design system written in code. We are migrating our last Python services to TypeScript over the coming months. WHO DOES WELL HERE
• You have built something that mattered: an early engineer at a startup that took off, a founding engineer, a product you launched with real users. You can tell us who it was for, what problem it solved and which number moved.
• You have turned a manual process into a product before, or you really want to.
• 5+ years building full-stack products in production, mostly in startups or fast growing companies where you had to decide with incomplete information.
• Strong in TypeScript, backend and frontend. A real eye for interface quality and the discipline to follow a design system.
• Solid engineering habits: system design, clean refactoring and a good nose for the risks of generated code that nobody tested.
• You can say no, reduce the scope and explain a technical decision to a founder in two minutes. WHAT WE EXPECT FROM YOU First week
• Get onboarded on the stack.
• Ship your first improvement on the dashboard. First month
• Own one feature end to end, from the problem with the ops team and the clinics to production, and show what it changed.
• Write down how the platform should evolve. First 3 months
• The first version of the scenario engine is in production.
• A standard clinic can be onboarded without a developer ticket. 6 to 12 months
• Most new clinics are live within 7 days.
• Clinics configure their own rules in the dashboard.
• You have hired at least one engineer and they ship on their own. PROCESS: 4 STEPS, FAST
• 45 min with Nans (Talent).
• 30 min with Hugo (CTO) or Pierrick (Head of Engineering).
• 2h in our office: live coding without AI, a design system exercise and a product case with Charles, our CPO.
• 30 min with the one you have not met yet. Then references and an offer. When the match is obvious, we have made an offer in 24 hours. You get feedback at every step. PRACTICAL
• Paris, rue d'Alexandrie (Sentier), new offices.
• 70-100K + BSPCE.
• Hybrid 3 days at the office.
• Claude Max provided.