In 2010, we started Unicorn as a software development agency.
For a decade, we built products for other people. Startups, enterprises, founders with ideas and budgets. We were good at it. We got fast at it. And somewhere along the way, we noticed a pattern that we couldn’t unsee.
Every project started the same way.
Database setup. Auth configuration. API gateway decisions. Workflow logic. Environment plumbing. Security layers. The same foundational decisions, made from scratch, every single time — for every client, every product, every engagement.
We were some of the best builders we knew. And we were spending 30–40% of every project just laying the same foundation we’d already laid a hundred times before.
In 2020, we made a deliberate decision: stop building for other people. Start building for ourselves.
We became a product studio. And then we started asking a harder question.
The question that changed everything
We looked at the best venture studios in the world — the ones that launch company after company with unusual speed and consistency — and we reverse-engineered what made them work.
The answer wasn’t just talent. It wasn’t just capital. It was shared infrastructure.
The best studios aren’t starting from zero with each new venture. They’re running the same playbook on a proven foundation. Same architecture. Same patterns. Same team. They’re compounding — each new company making the system stronger rather than starting from scratch.
We had spent 10 years building that foundation for other people.
It was time to build it for ourselves.
Scaffald
In 2024, Unicorn became a startup incubator. And the first thing we built wasn’t a product.
It was the infrastructure every product would run on.
We called it Scaffald.
The stack: Supabase for the backend core (database, auth, storage, realtime), Kong as the API gateway and security layer, Temporal for durable workflow orchestration, and an AI layer running across all of it to assist in development and automation.
But Scaffald isn’t just a tech stack. It’s the operating methodology for how we build companies.
When Unicorn launches a new venture, we’re not assembling a new team and starting over. We’re deploying the same core team — mostly agency alumni who’ve been building together for years — onto the same proven infrastructure. The founders of each Unicorn venture are the Unicorn founders. The stack they launch on is Scaffald.
That means when we go from 0 to 1, we do it with:
- A team that has already built together
- Infrastructure that’s already proven
- Patterns that are already standardized
- Workflows that are already designed
The goal: launch production-grade companies in weeks, not months. Reduce the infrastructure tax that kills early momentum. Let founders focus on the actual problem they’re solving.
Why build this in public?
Because we think the model matters as much as the technology.
There are a lot of people building venture studios right now. There are a lot of people building developer infrastructure. Very few are doing both deliberately, as a unified system, with shared IP at the center.
We’re going to document the whole thing — the architecture decisions, the tradeoffs, the workflows, the mistakes, and where this is going.
If you’re a founder, a builder, or someone thinking about how to build companies more systematically — this series is for you.
Next post: the technical architecture of Scaffald — why we chose Temporal for orchestration when most teams would just reach for a queue, and what that decision unlocks.
Follow along. We’re just getting started.