From finance and strategy into product and software.
I am a finance-trained builder from the Netherlands. Strategy and operations work at Peakable led to co-founding Datagate, and I now build independently on publishing software and AI research.
Background
I studied International Business with a Finance major at Maastricht University and spent one exchange semester at City University of Hong Kong. The lasting part of the training was learning to weigh trade-offs, downside and uncertainty quickly.
At Peakable I worked across strategy and operations for more than 20 clients: growth questions, financial models and several digital-asset projects. It showed me how often an attractive plan differs from an operating reality.
I co-founded Datagate, an AI-powered social-media analytics venture, where I ran strategy and product. We built an MVP in ten weeks with a team of nine and pitched more than 20 investors. Datagate never became a lasting company, and the case study is candid about why.
Now I build independently. LokiOS is a private, self-hosted publishing system; alongside it I am researching whether firm-level AI exposure can partially hedge human-capital risk, and testing what local models make possible for one-person businesses.
Experience
Maastricht University
Finance major across the International Business programme: trade-offs, downside and uncertainty.
Peakable
Strategy and operations across more than 20 clients: growth, financial modelling and operating questions, including digital-asset projects.
Datagate
Defined the product, coordinated a nine-person team through a ten-week MVP build, developed the financial and go-to-market models and pitched more than 20 investors.
City University of Hong Kong
One exchange semester within the International Business programme.
Independent work
Building LokiOS, running the human-capital research and testing local-model tooling for small businesses.
How I work
Start with the constraint
On the human-capital research, most of the early effort went into the occupation-to-firm mapping, because every downstream number depends on it. Naming the binding constraint first is cheaper than optimizing around the wrong one.
Make uncertainty visible
LokiOS keeps drafted claims separate from verified ones, and the research keeps exposure measures separate from employment outcomes. When a number is inferred rather than measured, the workflow says so.
Verify inside the workflow
LokiOS checks a draft against the live WordPress theme before publishing rather than after, and returns a structured report when a check fails. A check that runs outside the workflow protects nothing.
Treat the business model as product design
At Datagate, the financial model changed product decisions weekly: pricing assumptions altered what the MVP had to demonstrate. Monetization is a design input, not a layer added later.
Working interests
Publishing systems
LokiOS takes an article from evidence to published page.
Local AI infrastructure
What local models make possible for one-person businesses.
Software products
Small products and online businesses, priced and distributed as part of the build.
How AI changes work
Whether firm-level AI exposure can partially hedge human-capital risk.
I am based in the Netherlands, building LokiOS and running the human-capital research alongside it.