What we stand for. What we refuse.
There are three of us. We build. We own it. We sign it. This text is not a pitch. It’s a charter — and a line we refuse to cross.
The Salesforce ecosystem deserves better.
Twenty years deploying Salesforce. Twenty years watching talented operators hack together solutions around a tool that promises a lot and delivers slowly. Twenty years of unanswered IdeaExchange tickets, Winter ’XX releases without the requested features, Dreamforce keynotes that ship three years later — when they ship at all.
The ecosystem organized itself around the gaps. Copado for enterprise DevOps, Provar for testing, Gearset for release management. Serious people built good things. But white spaces remain. Places where the pain is real, daily, and where nothing exists.
Dopamine Labs exists to fill those spaces. Not all of them. The ones we know, the ones we’ve lived, the ones where we can ship something solid without raising 50 million.
What we refuse.
- AI bullshit. We code with AI, we own it — and we don’t sell “powered by GenAI” the way others sold “powered by blockchain” in 2018. AI is an internal tool. The customer buys a result.
- 48-row pricing matrices. One price, one value, one page. If we can’t explain it in 15 seconds, we’re hiding behind complexity.
- Demos that lie. No Photoshop mockups on the landing page. The terminal, the logs, the traces we show — they run for real.
- The roadmap-as-state-secret. Our roadmap is public. Our issues are public. Our RFCs are public. If a competitor learns what we’re building, good — it forces us to execute well.
- Vague sales. No opaque “contact us”. No qualification call before seeing the product. We’re available, we respond, we know what we sell.
What we stand for.
- Open-core. The engine is open, auditable, forkable. The business lives in integration, support, and governance. If a customer wants to self-host, they can. If a competitor wants to fork, go ahead — we’ll keep the execution advantage.
- Vibe coding, owned. Humans in the loop, AI at the keyboard. That’s our structural advantage in 2026. Not a secret, a choice. Three people produce what used to take thirty ten years ago.
- Signed changelog. Every release is signed by the person who shipped it. No anonymous “release notes”. A human decided, a human shipped, a human answers.
- Design partners first. The first five customers don’t pay. They shape the roadmap. We listen more than we talk for the first six months.
- France as a base, not a limit. We’re French, we vibe code in French internally, we document in English when it’s for the world. No inferiority complex. No posturing either.
The link with Dopamine Technologies.
Dopamine Technologies is the parent company. Salesforce consulting since 2005. Twenty years in the field, on projects, on tickets that don’t forgive. Stéphane Hermet learned there that shipping a software product is not a consultancy’s job. That it required a different DNA, a different structure, a different economics.
Dopamine Labs is therefore a separate entity. Same family, two crafts. Tech sells human consulting; Labs sells standard software. No forced cross-sell: if you’re a Tech client, nothing obliges you to buy Labs. If you’re a Labs client, nothing obliges you to go through Tech. The value must stand on its own, on both sides.
What binds the two: standards. No product compromise because consulting revenue is coming in. No sloppy consulting because the product is sexier. One red line: we don’t sell what we wouldn’t use ourselves.
Why now.
In 2020, shipping a SaaS required a twenty-person team, a year of development, and two million in pre-launch funding. In 2026, a three-person team with the right tools (LLMs, Next.js, Supabase, Stripe, Resend) ships in months what used to take a year. The marginal cost of a niche product dropped by an order of magnitude.
What was economically impossible five years ago is viable today. A product serving a thousand precise accounts, at a fair price, built by three people — that’s a healthy business in 2026. That’s what we’re building.
Version 1 · April 2026 · To be updated when our practices evolve — with a revision note.