Tippin' Monk
A fintech app that connects to 600+ EU banks to change how you think about spending
Most personal finance apps tell you what you already know: you spent too much on food last month. Great. Now what?
I wanted something that actually changes behavior. Not through guilt or budget caps, but through experiments. What happens if you cook three more meals this week? What if you switch your commute for a month? Small, testable changes with real data behind them.
I’d been thinking about this since my MBA. The research is clear: people don’t change spending habits because of pie charts. They change when they have a concrete experiment to try and can see the results in real time.
How it works
Connect your bank, see your spending patterns, pick an experiment, track the outcome.
Under the hood: Tink API for bank connections (600+ EU banks), a custom recommendation algorithm that suggests experiments based on your spending, and Claude + n8n to make sense of messy bank transaction descriptions. React + TypeScript frontend on Vercel, Supabase for auth and data.
The banking data problem
Building with real banking data is humbling. Transaction descriptions are chaos. Every bank formats them differently, abbreviations are inconsistent, and merchant names are rarely clean. The AI categorization layer took as much effort as the rest of the app combined.
Where it’s at
This is a side project, so “impact” is personal:
- Built and shipped a production fintech app end-to-end, solo
- Integrated with real banking APIs handling real money data
- Currently live and usable
- Learned more about product thinking from this one project than from any course
The other big lesson: scope creep is real when you’re building for yourself. I had to learn to ship the thing I needed, not the thing I imagined.