DependaPou

Inspiration

Research shows that 42% of developers spend most of their time dealing with technical debt. On average, developers dedicate 17.3 hours per week to maintenance tasks such as debugging and refactoring. This ongoing struggle with complex, over-engineered, or low-quality code can be mentally exhausting, contributing to developer burnout.

Inspired by this challenge, and the power of Large Language Models (LLMs) to automate workflows, we built DependaPou, a developer tool designed to offload much of the maintenance process to Generative AI.

What It Does

When you upload your GitHub repository to our dashboard, DependaPou:

Scans your codebase for outdated dependencies and vulnerabilities using the Groq Inference API running in a Modal container.
Generates a list of files that require refactoring.
Parallel processes each file using additional Modal containers with Groq's LLM to:
- Refactor the code.
- Provide clear, AI-generated comments.
Bundles the changes into a new GitHub branch, where they are:
- Staged, committed, and pushed.
- Opened as a pull request (PR) for review.
Notifies users via a Next.js dashboard, with updates stored in Supabase.