About me
From a 13-year-old with a Quake clan site to AI advisor
I was 13 or 14 years old when I first put a website online. Not because I was particularly fascinated by websites as such, but because I wanted to build a clan site for the group of friends I played Quake with.
When that site went live for the first time, I felt something I can barely put into words. Pride, yes. But mostly a kind of fulfilment. The feeling of: I had an idea, and now it actually exists. I've been chasing that feeling ever since.
The start of Pixelpillow
On 1 January 2013, I co-founded Pixelpillow with my partners. For nearly twelve years I built that company up — from a small studio to an organisation with multiple people, clients, and responsibilities.
It was a period of growing, learning, and above all making things. Websites, apps, brands. For clients who trusted us to bring their digital vision to life.
Jasper and the art of asking the right questions
My first real aha moment came in January 2021. Jasper.AI launched — a tool that could write entire blog posts based on your input. At the time I was mainly focused on content for Pixelpillow.
What struck me most was how Jasper interviewed me. What's your target audience? What tone? What are the key points? And that made all the difference. Because by asking the right questions, Jasper got better input from me. And better input produced better output.
ChatGPT and the world opening up
I still remember exactly what my first prompt was. Not because it was such a brilliant question, but precisely because it was so ordinary: "What's the best way to negotiate buying the garage from my neighbour?"
ChatGPT gave me a nuanced answer. Tips on how to approach that conversation. How to gauge what a fair price would be. That was the moment I realised: this isn't just a better version of Jasper. This is fundamentally different. This is having a conversation. This is thinking things through together.
The first Chrome extension in 3 hours
That came when vibe coding became a thing. I still remember my first project clearly: a Chrome extension that let me label reactions on my LinkedIn posts with colours.
I opened Cursor. I typed what I wanted. And an hour later I had a working first version. Two or three more hours of fine-tuning, and I had a Chrome extension that did exactly what I wanted. Three to four hours in total. From nothing to working.
That Chrome extension was the moment I realised: this isn't an evolution, this is a revolution. Because if I can build something in three hours that would otherwise take me a day and a half, that doesn't mean I'm four times faster. It means I can try four times as many things.
Saying goodbye to Pixelpillow
On 1 November 2024, I signed the transfer documents. The company was no longer mine. Honestly? It didn't come out of nowhere. I'd been wrestling with a question for a while: what's our right to exist as an agency, now that AI is changing everything?
I could see how fast things were moving around me. How AI tools kept getting better. How with vibe coding you could build in days what used to take weeks. And I wondered: if a freelancer with Cursor can build something in a week, why would a client hire an agency that takes months and charges ten times as much?
When I think now about how we would have needed to change our agency to keep up with all the developments, it would have required almost a complete turnaround. Not a little adjustment, but truly reinventing who we are and what we do.
Where I am now
After the sale, I deliberately decided to do nothing for the first six months. No commitments. No pressure. Just experimenting with vibe coding. As a hobby. For myself. With no obligation for it to lead anywhere.
But over the summer, that changed. The risk of immediately trying to launch a new product in this dynamic period suddenly didn't seem like such a good idea. Because there's a good chance one of the big tech companies in America is developing the same idea. And then you're wiped off the board within six months.
So my thinking became: I'll ride out the AI storm. In the meantime, I'll do advisory work. I stay sharp, I earn money, I help companies with their AI challenges. And in a year or two, when things have settled a bit, I'll see what kind of company I want to start.
And above all: I now have the mental bandwidth I wouldn't have had if I still had Pixelpillow. That freedom to really experiment. To try new things. To work out four different ideas a month just because you enjoy it.
Why I help companies with AI
Every day I see organisations struggling with the same question: where do you start with AI? What are the opportunities? And how do you make sure it goes beyond an experiment and creates real impact?
My approach is no-nonsense and practical. No theoretical frameworks or thick reports that end up in a drawer. I've applied this process for years at my agency. Now I help other organisations with it.