Import your Chime bank statements in PDF format, then chat about them with AI.
Note: This is a hacky personal project. Don't rely on this for important things...
- Log into your Chime account and navigate to your Documents: https://app.chime.com/settings/documents
- Click on the
PDF
link on the right. - Download every statement you can!
Place all your statements in a single folder and name them like:
Your_Name_Checking_eStatement (1).pdf
Your_Name_Checking_eStatement (2).pdf
The importer parses your Chime statements into a SQLite database for
analysis. For details on setting up and running the importer,
see the README in the importer
folder.
This app allows you to interact with your imported Chime transactions
using AI. Once you've set up your transactions.db
using the importer:
-
Start the AI Interface (Coming Soon):
./chime-ai
-
Ask Questions About Your Transactions:
- "How much did I spend on dining last month?"
- "What are my recurring subscriptions?"
- "Show me my largest transactions in the past year."
Prompt: Where do I spend most of my money? Give me the top 10 places in October
Response:
Based on the provided transaction information for October,
here are the top 10 places where you spent the most money:
1. **Transfer to Credit Builder**: $1,167.47
2. **Transfer to Chime Savings Account**: $1,166.32
3. **Foursight**: $631.02
4. **Transfer to Chase**: $525.00
5. **Progressive**: $361.38
6. **Clear Me**: $259.00
7. **Verizon**: $245.47
8. **Amazon**: $171.93
9. **Les Schwab Tire Centers**: $157.00
10. **Alaska Air**: $153.99
Your highest spending was on transfers and savings, which
constitutes financial management actions rather than direct
purchasing of goods or services from typical vendors or service
providers; this is something to keep in mind when analyzing your
expenses critically for budget adjustments or identifying areas
where you could potentially reduce spending if necessary.
- Import Chime bank statements into SQLite
- Clean up parsed text files
- Implement DB repository functions
- Add AI chat functionality for transaction queries
- Add tool calling and open ended query building agent
- Improve error handling and data validation
- Support for additional bank formats