Logo
Homepage
Demo
Demo
Demo
Demo
Demo
Demo

RepairALL: AI-Powered Open Source Guide for Everyday Repairs

Team Members

Louis
Jules

Inspiration

We were inspired by the idea that knowledge should be accessible to everyone, especially when it comes to fixing everyday objects. Millions of perfectly usable items are thrown away each year simply because people don’t know how to repair them.

RepairALL aims to change that.

Our goal was to create an open source platform that teaches anyone, anywhere, how to repair what they own, through a friendly AI assistant that generates visual, step-by-step repair tutorials from a simple photo and short description.

This way, knowledge about repairs becomes a shared resource, not a privilege.

What It Does

RepairALL uses AI and image generation to transform a broken object into a personalized repair guide.

Here’s how it works:

The user takes or uploads a photo of a broken object.
They describe the problem in a few words (e.g., “The leg of my chair is cracked”).
RepairALL’s AI — powered by Gemini API — analyzes the photo and text to understand the issue.
It generates 3 to 5 tutorial images with text instructions explaining how to fix the object safely and efficiently.

The result: a personalized, accessible repair tutorial, making DIY learning easy for everyone.

How We Built It

A responsive web application built with React and TypeScript. It allows users to upload images and view the generated repair guides directly in their browser.
AI Reasoning: Google Gemini API analyzes the image and description, generating tool lists, repair steps, and safety notes.
Image Generation: Gemini API creates realistic images showing each step of the repair.
Styling: Tailwind CSS for a clean, responsive interface.
Authentication: Auth0 for secure and seamless user login.
(Hosting information removed as requested)

All our code, documentation, and model prompts are openly available on GitHub under an MIT License, making it easy for others to contribute, learn, or improve the project.

Challenges We Faced

HEIC image conversion: Handling iPhone photo formats without breaking uploads.
AI reliability: Ensuring Gemini produces accurate, safe repair instructions.
Latency: Coordinating multiple API calls (text + images) within seconds.
Designing for inclusivity: Building a UI accessible even to users unfamiliar with tech or DIY repair.

Every challenge helped us refine the system to be faster, safer, and more community-friendly.

Accomplishments We’re Proud Of

We built an AI-powered repair assistant fully functional in under 48 hours.
We successfully integrated Gemini to combine reasoning and visual generation.
The entire project is open source — anyone can fork, reuse, or contribute.
Our app empowers users to reduce waste and learn through repair, embodying the true spirit of open knowledge.

What We Learned

AI for good: Combining technology with sustainability creates meaningful social impact.
Prompt design: Fine-tuning how we communicate with models like Gemini is key to useful results.
Open source collaboration: Writing clear documentation and licensing empowers others to build upon our work. We also realized that making knowledge open-source is not just a technical decision — it’s a cultural one.

What’s Next for RepairALL

Launching a community repair hub, where users can share their generated guides and improve them collaboratively.
Adding multilingual support so that repair tutorials are accessible worldwide.
Integrating voice and AR features for real-time, hands-hands-free repair guidance.
Creating a public dataset of repairs to help other open source developers train sustainable AI models.