HappyGF: I Built an AI Assistant to Make Sure Contractors Get Paid on Friday
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle . What I Built Every Friday, the accountant processes payments to the contractors and suppliers who keep my girlfriend's business running. For that to happen, invoices need to be received, reviewed, approved, correctly named, and forwarded to the accountant before the deadline. It sounds straightforward, until…
This article discusses the creation of an AI-powered accounts payable assistant called HappyGF, designed to simplify the process of processing invoices for the girlfriend of the creator, who runs a business. The original system of processing payments, invoicing, and forwarding documents was time-consuming and prone to errors, often resulting in missed payments to contractors and suppliers.
The automated assistant, HappyGF, was built using Java and an open-source framework called The Pipeline Framework (TPF). The AI workflow utilizes two open-weight models: Gemma for understanding invoices and interpreting text and visual information, and Tev1 for structured decision-making. Gemma handles the interpretation of invoices, which often come in various formats and contain different information.
Tev1 evaluates predefined questions against the invoice information and returns structured decisions, allowing the assistant to correctly name PDF files, track processing state, detect duplicates, and support the approval workflow.
The application uses a Microsoft Graph connector for Outlook and implements business rules and authorization under its control, rather than relying entirely on AI decisions. During development, the creator used Ollama for local model inference and automated integration testing. HappyGF is now open-source, allowing other businesses to adapt and utilize the application for their accounts payable workflows without relying on proprietary AI services.
This case highlights the benefits of open innovation, particularly in sensitive business areas like accounts payable, where businesses may prefer to keep sensitive information on their own hardware. By utilizing open-weight models like Gemma and Tev1, this assistant demonstrates the advantages of flexibility, cost-effectiveness, and maintainability compared to proprietary solutions.
The creator's girlfriend has found the application valuable and is using it for her actual accounts payable work, highlighting its real-world utility.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
