Equity crowdfunding platform
A startup needed its first product, and I was the only developer. I built it end to end and shipped the first version in 7 months.
Full-stack engineer and engineering leader. I'm a Director of Engineering. I also founded Turbosoft, where I'm building Garnet. It is a private, local-first notebook for notes, a journal and everyday records.
A startup needed its first product, and I was the only developer. I built it end to end and shipped the first version in 7 months.
I joined a fast-growing Copenhagen-based hardware startup when it was about 20 people, and left when it was close to 100. I started as a frontend developer on the online store. It later moved from WordPress to Shopify and Next.js, and served 2M+ visitors a month. The phone app controls the product over BLE. It already had its direction when I moved onto it, so I built on it with the team. To tie them together, I helped build a cross-platform UI library, so the store and the app shared the same components. Both ran on one end-to-end stack, from the interface to the backend services.
A company ran a main product, experiments, internal tools and a blog. Each had its own team. The main product carried years of technical debt, and few people could maintain it. I was brought in to untangle it.
I led the team through a long-term plan. The goal was to pay down the debt and to experiment without hurting the main product. We also brought data people, designers and developers into one process. I made the architecture calls and stayed hands-on in the code.
It was by far my largest long-term project, and it paid off. Three years in, microservices and micro frontends let us move people between products easily. We tried new ideas much faster, with far less effort.
I built a headless e-commerce site for a leading premium food and wine brand in Denmark. It covers the product catalogue, basket and checkout, and wine club membership. Recipes and a store finder sit next to the shop.
Alongside my full-time roles, I ran a small consulting studio. There, I helped a 20+ person team set up AI chat on their own hardware. We focused on privacy, operability and long-term cost.
One of the platforms I built was for a major Danish trade union and unemployment insurance fund. Each component maps to a block in the CMS, so editors put pages together themselves. The CMS is used by 20+ editors and powers the content for 2,000+ pages.
Then I founded Turbosoft. I'm building Garnet there, end to end.
People do their best work when they have a say in how they work, get better at it, and know why it matters. Give them all three.
Feedback should be kind and clear at the same time. Staying quiet to be nice helps no one, so say it early and with care.
Share the goal and the why, and step in only when something is not right. When two options are equally good, the team's choice wins.
A team can only hold so much in its head. Size what each team owns to what it can carry, and cut accidental complexity first.
Effort is not the goal. Put energy where it moves the product, not where it is easiest to look busy.
AI helps teams ship faster and try more ideas. People stay responsible for the quality of what they ship.
A leader is measured by what the team ships, not by their own work. So spend time where it lifts the team the most.