Pesarifu

Pesarifu is an application for conducting data analysis of financial statements, primarily from mobile transactions. Current features include:

• Data Extraction from PDF Statements into SQL Tables. Supports PDFs from
  • Safaricom (MPesa)
  • Standard Chartered
  • Equity
• Export transactions to CSV, JSON and Excel files.
• Report generation using Evidence

See examples/mpesa.pdf for an example statement and examples/Financial Report by Pesarifu for John Mwangi.pdf for an example output report produced by the application.

Additional documentation can be found in docs/.

Usage

Setup

• clone and cd into repo
• set your own values for the variables in .env.example and load it into the environment source .env.example
• ensure the following are defined in pesarifu.config folder
  • In settings.toml
    • CELERY_BROKER_URL - URL for celery broker, expects Redis
    • CELERY_RESULTS_BACKEND - URL for celery results backend, expects Redis can be same as CELERY_BROKER_URL
    • APP_BASE_URL - base web URL that app is deployed
    • STATEMENTS_BASE_DIR - where financial statements are stored after upload
    • EXPORTS_BASE_DIR - where csv/excel etc exports are stored before being sent
  • In .secrets.toml
    • MAIL_USER - email used to send report to user.
    • MAIL_PASSWORD - password for the MAIL_USER email
    • TELEGRAM_BOT_TOKEN - token for Telegram notifications
    • DB_URL - production and development url for accessing DB through SQLAlchemy, expects postgres e.g DB_URL = "postgresql+psycopg2://postgres:@localhost:5432/pesarifu"
• run poetry install
• run npm install
• run alembic to setup the database alembic upgrade head
• cd into reports/
  • run npm install
• (optional) run python -m pesarifu.db.seed to seed the database with some example data for a report.

Running

• Make sure redis and postgres are running
• Use just celery-run to start celery workers
• Use just app-run to start the web app
• Use just reports-serve to start the evidence report back-end