Cluj-Napoca · remote EU or hybrid · contract or permanent
Usually that's a migration nobody can afford to let stall, or an API that got slow and nobody's sure where. Sometimes it's a regulated backend where the edge cases are the actual product. Most recently that was an institutional crypto-custody platform for banks at Bitpanda, and before it ImmoScout24, RWE, and a run of systems that had to keep working while they were rebuilt underneath.
I ask a lot of questions in the first week and considerably fewer after that. Then I take the difficult thing off your plate, and it ships tested.
The best fit is usually a problem that's been sitting on the board a while: awkward, a little risky, never quite this quarter's priority. That's the work I want.
Bank-facing services where a wrong answer costs somebody real money. I owned the crypto address book at Bitpanda, and in custody a bad address quietly accepted means funds nobody gets back. So I built it backwards from what must never happen.
Evidence Bitpanda — institutional crypto-custody for banks.
Old service, small team, no appetite for a rewrite that stops everything. At RWE we moved it across in slices with tests around each one, so delivery kept going while the ground shifted underneath.
Evidence RWE — Java to NestJS on an energy-market platform.
I measure before I touch anything. Most slow systems turn out to be slow in one place nobody has looked at yet, and a week spent guessing is a week gone. The export figure below came from profiling first and fixing what the profile actually showed.
Evidence About 65% faster APIs. Exports from 45 minutes down to 12.
Joining a team that has its own conventions and becoming useful inside them quickly. No rewrite proposals in week one. I'd rather learn why it was built that way first.
Evidence ImmoScout24 — search, listing, SEO, Iterable.
At 16 I won an internship contest in Galați and turned it into my first software job, doing Linux and web work on things people actually used. Computer Science at Babeș-Bolyai came after, and mostly it put names to habits I'd picked up by breaking things in production and having to fix them.
Fintech and custody, energy, real estate, publishing. A DAX-listed energy company, a custody platform for banks, one of Austria's biggest property portals. The pattern is usually the same. I get handed the part everyone has been working around, and I stay on it until it's genuinely done rather than technically closed.
Remote EU or Cluj, contract or permanent. The shape matters less than whether the problems are hard enough to be worth doing properly. What I want is a product or platform team with real domain complexity, where people talk to each other and own what they build.
Four systems, what was actually wrong with each, and what changed.
Owned the crypto address-book backend end-to-end in a regulated custody platform for banks, strengthened auth across services, and introduced testing patterns the team adopted.
Shipped search/listing, SEO, and Iterable integration work inside a fully tested distributed platform where release quality and production safety mattered.
Reduced delivery risk in a Java-to-NestJS migration, profiled MongoDB-heavy data flows, and moved export processing into worker pools.
Rotated across enterprise domains, shipping React and Node.js systems for payments, telecom, and publishing clients with fast ramp-up and clear delivery habits.