We kept rebuilding the same spreadsheet. So we stopped.

Léa spent six years running treasury at a French payments company, watching the same failure mode repeat itself every quarter: a forecast built carefully on a Monday, quietly going stale by Wednesday, and nobody noticing until a tight week arrived without warning. Marco had spent almost as long building time-series forecasting systems for a logistics platform, solving a version of the same problem for delivery volumes instead of cash.

They met in 2023, compared notes, and realized the tooling gap wasn’t a lack of forecasting technique — treasurers have used rolling 13-week forecasts for decades — it was that nobody had built a system that kept one continuously updated from live bank and accounting data, the way a logistics platform keeps a delivery forecast updated from live tracking data. Dibein started as a weekend prototype connecting one bank account to a spreadsheet-replacement script, and became a company once three friends running finance at their own startups asked to use it for real.

We’re a small team, based in Rueil-Malmaison, just outside Paris, close enough to La Défense that most of our early customers were a short train ride away for the first round of in-person onboarding calls. We stayed EU-based and EU-hosted deliberately — the regulatory bar for handling financial data in Europe is high, and we’d rather build to it from day one than retrofit it later.