Statement Sensei helps converts bank statement PDFs to CSVs using the monopoly CLI library. The offline version of the app is available on the releases page.

🎉 Statement Sensei is now live! 🎉

Try it out:
https://statementsensei.streamlit.app/

Usage

Statement Sensei can be run as an offline application on Windows, MacOS or Linux.

The offline application runs Streamlit locally, and uses a WebView window to view the browser frontend at http://localhost:8501.

Currently supported banks:

Bank	Credit Statement	Debit Statement
Bank of America	✅	❌
Chase	✅	❌
Citibank	✅	❌
DBS/POSB	✅	✅
HSBC	✅	❌
Maybank	✅	✅
OCBC	✅	✅
Standard Chartered	✅	❌
UOB	✅	✅
Zürcher Kantonalbank	❌	✅

Installation

Warning

The offline app may raise security warnings during installation.

Specifically on MacOS, the application will show an "app is damaged and can't be opened" error.

These security warnings happen because the release binaries are unsigned, and are incorrectly flagged as malware.

To get around this, follow these steps for MacOS / Windows.

The Windows Defender alert can be bypassed by clicking "More info" -> "Run anyway".

Development

Install system dependencies using brew or apt-get (necessary since pdftotext needs them)

apt-get install build-essential libpoppler-cpp-dev pkg-config ocrmypdf

or

brew install gcc@11 pkg-config poppler ocrmypdf

Install app dependencies with Poetry:

poetry install
poetry shell

To run the consumer-facing application:

python entrypoint.py

To run the application in developer mode:

streamlit run webapp/app.py

Docker

Otherwise, to run the application as a container:

docker compose up

or:

docker pull benjaminawd/statementsensei:latest
docker run -p 8501:8501 benjaminawd/statementsensei:latest

If running locally with docker: you can either store passwords in an environment variable as a string

export PDF_PASSWORDS='["pass123", "otherpw123"]'

or store them in an .env file in the project root:

echo 'PDF_PASSWORDS=["foo"]' > .env

Features

• Supports uploading multiple bank statements
• Allows unlocking of PDFs using user-provided credentials via the frontend