Automating Invoice Processing with GPT-4o

This repository contains the code for an automated invoice processing system using Python, OpenAI's GPT-4o and Pandas. The system can process multiple PDF invoices, extract relevant information, perform basic financial analysis, and generate an Excel report.

For a detailed explanation of the code and concepts, check out this blog post.

Structure

The code is organized into the following files:

• main.py: The main script for the invoice processing pipeline.
• extraction.py: Extracts the data from PDF invoices using GPT-4o.
• processing.py: Validates the invoice data, performs the financial analysis and creates the Excel report.
• config.py: Configuration settings for the application.
• prompt.py: Defines the system prompt used for GPT-4o.

Installation

Prerequisites ✅

• Python 3.11 or higher
• Poetry (Python package manager)

Instructions

1. Install the dependencies using Poetry:

   poetry install

2. Create a .env file by copying the provided .env.example file and set the required environment variable:

   • OPENAI_API_KEY: Your OpenAI API key.

Running the Application

Before running the application:

1. Update the COMPANY_NAME in app/config.py to match your company's name. This is used for invoice type classification (incoming or outgoing).

2. Ensure your PDF invoices are placed in the directory specified by INVOICES_DIR in app/config.py (default is data/invoices).

To run the invoice processing application:

poetry run process-invoices