I'm a Dutch guy living in Singapore. By day, I lead Trust, Safety and Customer Experience at Carousell, Southeast Asia's largest classifieds platform. By night (well, more like early mornings before the kids wake up), I build things with code. Most recently Magicdoor.ai, a ChatGPT alternative I shipped entirely by myself.

That second part would have been impossible two years ago. I couldn't code. Now I ship full-stack products. This transformation, and what it means for how we'll all work in the future, is something I think about a lot.

The pivotal moment

In 2009, I was studying Biomedical Engineering in Groningen and pretty much failing at it. Then I read a newspaper article about the Global Financial Crisis. The headline was something like "Trillions wiped out with derivatives." I had no idea what any of that meant. CDOs squared and cubed, credit default swaps, all this financial alchemy that nearly broke the world.

I bought the newspaper, read the whole thing, and got completely nerd-sniped. I thought: I need to understand how this works. I switched tracks entirely, got my Masters in Finance, and started my career in Private Equity at AlpInvest Partners in Amsterdam.

The Excel years

I absolutely loved building financial models. I removed the F1 and Caps Lock keys from my keyboard. I could build a leveraged buyout model from scratch faster than my McKinsey colleagues could open Thinkcell. It was intellectually stimulating, like solving elegant puzzles.

But after a couple of years, I got bored. The work didn't really change. Your decisions don't get much better if you make more intricate models. And the career path seemed so predictable. Associate, then VP, then whatever comes next. I wanted more adventure.

Bangladesh, Myanmar, and why not today?

In 2015, I quit my job, and flew to Bangladesh for Rocket Internet. A friend there told me on a call: "You want to build Amazon in Myanmar? You can fly next Tuesday." I said I needed a bit more time (there was a bonus coming), but I knew then I'd found my people.

The speed was intoxicating. On my first day in Bangladesh, we needed internet. I figured out who had the best connection, made the call, and literally two hours later guys showed up with fiber coiled around their shoulders, absailed down our building, and routed it through a window that could no longer close. When I later started Ox Street in Singapore, it took five weeks to get internet.

At Rocket, I learned something I still use today: Why not today? When someone says "we can do this next Tuesday," I'll ask "okay, but why not today?" It's not meant to be aggressive. It's a genuine question. What actually stops us from just doing it right now?

I ended up running jobs marketplaces across Myanmar, Bangladesh, Cambodia, and Sri Lanka. Over 100 people eventually. I learned to build teams, do sales, market products. I also learned to DJ and helped start a techno club in Yangon called Level Two (it still exists, much bigger now).

The Bali detour and the wrong question

After Rocket, I went to Bali for what was supposed to be a sabbatical to "find my passion." I surfed. I built little apps with no-code tools. I did freelance consulting. I had some truly terrible business ideas surrounded by digital nomads who also had terrible business ideas.

I now think I was asking the wrong question. I was looking for some magical type of work that would make me obsess and lose track of time forever. But that's not how it works. I'd once obsessed over Excel models, then over building companies at Rocket, and the obsession always faded when things got easy or predictable.

The real answer, which I've written about on my Substack, is that meaning and purpose emerge from action, not contemplation. Passion often follows mastery, not the other way around. The craftsman mindset: focus on getting so good at something that you can't be ignored.

Ox Street: learning by doing

In 2019, I moved to Singapore and started Ox Street, a managed marketplace for sneakers.

We nearly went bankrupt twice. COVID hit just as we were running out of money. But then COVID saved us. No more meetups, people scared of germs, and suddenly authentication became essential. We grew 35% month-over-month for 10 straight months. That's when I actually saw a hockey stick in my actuals, not just projections.

I learned product management the hard way. Our designers were too expensive, so I taught myself Figma and did it myself. I learned to really work with engineers. I learned that the economics of regional marketplaces are brutal. You need global scale, not regional.

In 2021, Carousell acquired Ox Street. This was after a year-long relationship where we'd discussed partnerships. When their CEO asked if I'd consider selling, I actually didn't see it coming. The core authentication capabilities and operational knowhow lives on in Carousell Certified.

The enterprise chapter

Going from 30 people at Ox Street to 700 at Carousell was a culture shock. I'd never worked in a "big" company. There were budgeting processes, stakeholder management, politics. Some of it I looked at with wonder. Some of it I learned was actually necessary.

But the scale is motivating. In Customer Experience, we've dramatically reduced the human support burden through a combination of AI and fixing upstream user problems. In Trust & Safety, we've created a full order of magnitude reduction in phishing scams. When you're touching millions of users, the impact is real.

Learning to code

In 2024, I started learning to code with Cursor and Claude. I built Magicdoor.ai, a full Next.js app with authentication, payments, API integrations, the works. It has paying users. I use it every day.

The moment I realized what was possible: I needed some app store review data that the data team said would take two weeks. I found the API docs, went to ChatGPT, and said "write me a Google Apps Script to pull this data." Ten minutes later, data was flowing into my Google Sheet. I fixed one error by pasting it back to the AI. Done.

I'm now convinced we're at the beginning of a fundamental shift in how solo founders can operate. You can do so much more with so much less. Not "billion-dollar company with AI" less (yet?), but definitely "profitable side business with AI" less.

But I'm also realistic. Foundation models might be plateauing. Full-stack agents are still unreliable. And creativity, the ability to come up with the idea in the first place, that's still very much a human thing. AI is better thought of as an incredible unblocking tool rather than a replacement.

What I'd tell my 20-year-old self

Get into coding. 100%.

The irony is that when I started Ox Street in 2019, I decided not to learn to code because of no-code tools and the abundance of engineers. Now, in 2024, when AI can code, I'm finally learning. Because to work effectively with AI code assistants, you actually need to understand how code works.

I loved building Excel models for the intellectual puzzle. I loved building companies for the creative challenge and tangible results. Coding combines both. Infinite creative possibilities with something you can actually make money from.

Why this page exists

I wanted a place where someone could get a decent sense of who I am without having to piece it together from LinkedIn bullet points. If you're here because we're meeting for coffee, considering working together, or you're just curious about the guy building AI tools at 5am before his kids wake up, hopefully this helps.

And if you read this far: thanks. Feel free to reach out via LinkedIn.