Data analyst at P&V Insurance for four years. Building two live products that already serve real users — ARBITORIA and Salaire-Plus. Working between Korean, French, and English — a combination that's uncommon in Brussels and unusually useful for cross-border fintech and insurance.
I started at P&V Insurance as a data steward. For two years, my job was the cases the company's AI systems couldn't resolve — records that wouldn't merge, fields that disagreed across systems, edge cases the model handed back as uncertain. The work was unglamorous and slow. It was also the most useful education in data I could have asked for.
I learned that data quality isn't a problem you fix once. It's a continuous translation between what the business says it wants, what the systems actually capture, and what humans actually do. AI handles the patterns. What stays on a human desk is the texture — the cases that don't fit. That texture is where most real decisions live.
For the past two years I've worked as an analyst on the same data I once stewarded. The unexpected lesson: once you adapt to any role, the work starts to take on the same shape. Different domain, same pattern. Which is why I want the next thing to be different on purpose.
I want to choose what I do — not what I can do, not what I'm assigned. The work I want sits closer to building futures than predicting them.
Outside work I build two products that already serve real users. ARBITORIA covers Belgium and France — the financial decisions a household actually runs into, across salary, housing, utilities, insurance, loans, and investing, all on official data. About 300 people use it daily. Salaire-Plus is a focused salary transparency tool spanning six countries and six languages, built for people whose careers cross borders. On peak days it serves 3,000 visitors. Both exist because most financial decisions ordinary people make happen under information asymmetry that doesn't need to exist.
I work between three languages and two cultural contexts — Korean, French, English; Asian and European business norms. In Brussels, this combination is rare. It's the right shape for a specific kind of work: cross-border products in fintech and insurance, where regulatory nuance and language depth both matter — Korean companies coming into Europe, European companies looking at Asia, or anything in between.
A personal-finance transparency platform for Belgium and France. Salary, housing, utilities, insurance, loans, investing — built on official data only. Deliberately review-free: noise compromises transparency. The window through which a household can see their actual options against their own country's real numbers. About 300 people use it daily — verified through Cloudflare analytics.
Read case studyA focused salary transparency tool spanning six countries — Korea, Japan, U.S., France, Spain, Belgium. Gross-to-net calculation, country-specific job markets, cross-country comparison. For people whose careers and compensation conversations cross borders, in the language they're already thinking in. Peak days reach 3,000 visitors — verified through Cloudflare analytics.
Read case studySelected work from four years inside P&V Insurance — first as a data steward, now as an analyst. The steward years involved manually resolving cases the AI handed back as uncertain. The analyst years involve turning that better-quality data into business decisions. Specifics anonymized; methodology and outcomes described in detail.
Read case studiesEvery job, once you've adapted to it, takes on the same shape. The novelty wears off; the structure rises to the surface. The interesting question isn't what you can do — it's what you choose to want, given that almost anything becomes routine in the end.
The most useful thing my data-steward years taught me was what AI hands back. The cases the model can't decide are the cases where the data is most interesting — edges, ambiguities, exceptions that don't fit the pattern. Most real-world decisions live in that pile, not in the confident middle.
Prediction is the dominant frame in data — what will happen, how likely, how soon. It's a useful frame, but a passive one. The work I want next sits on the other side of the verb: not predicting what comes, but building it. ARBITORIA and Salaire-Plus are my smallest versions of that.
I'm self-taught. Four years at P&V, two live cross-border products with real users, three working languages. In data and software, this is no longer unusual — and the portfolio is why. The credential question matters less than whether the work is good.
If you're building cross-border fintech or insurance products — especially Korean companies expanding into Europe, European companies looking at Asia, or anyone whose product needs regulatory and linguistic nuance across markets — I'd like to hear from you. Equally interested in roles where the work leans toward building tools rather than producing reports.
Based in Brussels. Open to remote & contract roles globally — happy to be hired through EOR services like Deel or Remote.com, which most peer companies already use for EU candidates.