AI Issue Assistant: Turn Past Issues into Useful Insights

"AI Issue Assistant" is an innovative project that emerged from the Hacktoberfest Weekend Challenge: Build for a Friend. The tool is an AI-powered issue tracking system crafted to assist teams in managing and comprehending recurring issues more efficiently. Its inception was sparked by a common problem faced by teams: after encountering a particular issue multiple times, they often need to manually delve into past tickets to grasp the context of earlier occurrences.

The key features of "AI Issue Assistant" include:

- Streamlined issue creation and management via an intuitive interface

- Generation of vector embeddings to represent each issue

- Identification of similar historical issues using semantic similarity

- Simplification of spotting recurring problems

- Establishment of groundwork for future AI-driven functionalities like issue summaries and Root Cause Analysis (RCA)

By treating an issue tracker not just as a repository of tickets, but as a learning tool for the team's historical issues, "AI Issue Assistant" aims to make that valuable information more accessible and usable.

The project was developed using a modern full-stack JavaScript/TypeScript stack. The frontend was created with React TypeScript and Tailwind CSS, while the backend was built with Node.js Express.js and MongoDB Mongoose. The core AI functionality utilizes Gemini Embeddings to transform issue descriptions into numerical vector representations, which are then stored alongside the issue data.

These vectors facilitate the discovery of semantically similar issues, even when they employ different wording. For instance, "User cannot complete EPF authentication" and "EPF login is failing for a customer" might describe a similar problem despite differing phrasing. This capability of semantic search surpasses simple keyword searches.

The significance of open innovation is underscored in this project. It enables experimentation and learning to build useful AI features without heavy infrastructure or costly proprietary AI platforms. The project combined open-source technologies such as React, Node.js, Express, and MongoDB with accessible AI tools to create the complete application.

A crucial aspect for the developer was not merely incorporating an AI API, but understanding how AI features like embeddings and semantic similarity can address a genuine software development challenge. Open innovation also empowers developers to learn by building, modifying existing tools, and creating solutions tailored to their unique problems.