Aurelio
Aurelio is a wealth-management and financial-planning app that runs on your own computer, with an AI advisor you can talk to. Your records are kept on your machine. The AI part is optional: it goes through OpenRouter with your own key.
Try it on invented data
You need Git, Node.js 22 or later and uv. Then:
git clone https://github.com/LosaLosSantos/aurelio-finance.git cd aurelio-finance ./start.sh --demo # macOS and Linux
git clone https://github.com/LosaLosSantos/aurelio-finance.git cd aurelio-finance ./start.ps1 -Demo # Windows, in PowerShell
If PowerShell says that running scripts is disabled, use
powershell -ExecutionPolicy Bypass -File .\start.ps1 -Demo.
The first start installs what it needs inside the folder and opens http://localhost:8000 on an invented household, kept in a file of its own. Start without the flag to use your own data.
Run it on your own data
./start.sh # macOS and Linux./start.ps1 # Windows
Your records live in backend/data.db. Before an update changes its format,
the app saves a copy next to it.
What it does
- Wealth: your banks and brokers, their cash, and dated situations of what you hold.
- Portfolio: every position priced from the market, with its cost, gain and dividends.
- Look-through: what your funds really hold, by country, sector, company and currency.
- Plans, homes and debts: recurring investments, real assets and the loans that finance them.
- Cash flow and goals: income, expenses, savings rate, and the return each goal needs.
- Ask Aurelio: a chat that answers from your own records, can search the web, and proposes every change as a card you confirm.
- The analysis: an analyst and a confidant who knows you argue over your portfolio, and a synthesis says what to look at first.
How Aurelio reasons explains the ideas underneath. Guides, and how each number is worked out, are in the wiki: how to calculate net worth, ETF overlap and portfolio X-ray, accumulating vs distributing ETFs, how many months of emergency fund and what an AI financial advisor costs.
The AI part
- Create an account at https://openrouter.ai, add credit, and create a key at https://openrouter.ai/keys.
- Copy
backend/.env.exampletobackend/.envand setOPENROUTER_API_KEY.
The default model is Claude Opus 5.5 ($4 per million input tokens, $20 per million output tokens); you can pick another in the chat. What it cost, measured in October 2026 (the chat's rows on 11 and 12 October, an analysis on 12 October, a web search on 9 October, on a test database):
| cost | |
|---|---|
| A first question that drew a card, answered in one round | $0.11 to $0.12 |
| A follow-up, answered in one round | $0.021 |
| A follow-up that drew a card, answered in one round | $0.012 to $0.022 |
| Confirming a card, with the reply after it | $0.014 to $0.049 |
| Whether something is worth buying, and how much, in two or three rounds, its searches included | $0.067 to $0.24 |
| Each web search, on top of the round that asked for it | $0.0073 to $0.0075 |
| An analysis, its lookups included | $0.64 |
Development
uv run pytest in backend/ and npm test in frontend/. For hot reload,
run uv run uvicorn app.main:app --port 8000 in backend/ and npm run dev in
frontend/.
License
MIT, see LICENSE.