- Python
- Supabase
- Plotly
- Pandas
- Futu API
Finance OS — Independent
A personal financial dashboard built with the rigour of a small company's FP&A function. Net worth across asset classes, portfolio benchmarked against the right indices, cash flow at category level.
Premise
Most personal finance tools optimise for simplicity at the cost of insight. The aggregator apps tell you what you spent on coffee. They don't tell you whether your savings rate is consistent with your stated time horizon, or how your portfolio is performing against the right benchmark, or what your net worth would look like under three reasonable scenarios. Finance OS is built for someone who actually wants to know.
Approach
A Python pipeline ingests data from brokerage, bank, and CPF accounts via Futu's API, statement parsers, and direct entry where APIs don't exist. Everything lands in a Supabase store with a clean schema. A Plotly dashboard renders the views. Net worth curve by asset class, portfolio performance against benchmarks, cash flow by category with drill-down, and a scenario engine that runs three forward paths for any major decision.
The design philosophy is single source of truth. Every number on the dashboard traces back to a primary record. No estimates, no extrapolations, no plausible-sounding fiction.
Outcome
Six spreadsheets and three apps replaced by one system. The most valuable feature turned out to be the scenario engine. Being able to model a major purchase, a job change, or a portfolio reallocation against the actual numbers, rather than guess, has changed how I make medium-horizon decisions.