Youcanfixit

User uploads drywall damage photo and receives complete analysis: DIY cost savings, materials list, and step-by-step repair instructions
Cost Comparison Chart
Interactive floating tools: Click them and watch them spin away!

YouCanFixIt — Your AI Home Repair Companion

Everyone on our team has dealt with the same frustrating loop:

Something breaks at home.
You have no idea what the part is called.
You go down a 2-hour YouTube rabbit hole.
You still don't know if you're doing it safely.
You end up calling a professional and paying $200+ for a 10-minute fix.

Whether it was a leaking faucet, a jammed door lock, or a mysterious hole in the drywall, we realized people don't struggle because fixes are hard, they struggle because information is scattered, inconsistent, and often assumes prior knowledge. And when they give up and hire someone, they often pay 10x the cost of materials for labor.

So we set out to build an AI helper that makes home repair accessible for everyone, and saves them money in the process.

What It Does

YouCanFixIt is a visual troubleshooting chatbot for household repairs. Here’s how it works:

Upload an image of the problem (e.g., “my pipe is dripping,” “this hinge is broken”).
Add a short description of what’s happening.
Get a step-by-step, safe, beginner-friendly DIY repair guide instantly.

The chatbot breaks things down into:

The exact name of the part in the picture
What the issue is and why it happens
A repair plan with clear steps
Safety considerations (like how to turn off water or power)

How We Built It

We built YouCanFixIt using:

Frontend: JavaScript + React for a clean, responsive UI
Backend: Django + Python for API handling
AI: OpenAI GPT-4o-mini with custom prompt engineering
Plus: Caffeine + teamwork + determination

In short: A full-stack web app powered by AI and fueled by late nights.

Challenges We Ran Into

Time pressure: We had more ideas than hours, which forced us to prioritize core features over “nice-to-haves.”
Prompt consistency: Making the model give safe and correct repair instructions required constant refining.
Image interpretation quirks: Sometimes it thought a shower valve was a “weird metal mushroom.” That… needed fixing.
UI/UX balancing: We wanted a fun vibe without making the interface cluttered or confusing.

Accomplishments We’re Proud Of

Turning a shared frustration into something genuinely useful
Making a demo that actually works in multiple real-world scenarios
Keeping the hackathon fun, energetic, and collaborative
Building something that people immediately “get” when they see it
And honestly just seeing our idea go from “what if…” to “try uploading your picture!” was a huge win.

What We Learned

How to connect a frontend to a backend using React HTTP requests and Django API endpoints
How to structure prompts so the AI gives actionable steps rather than vague advice
The importance of UI/UX even in fast-paced hackathons
How to divide tasks cleanly and communicate effectively under time pressure
And of course: stay calm when npm breaks.

What’s Next for YouCanFixIt

More robust troubleshooting: Let users ask follow-up questions (“What if the screw is stripped?”).
Material recommendations: Auto-generate a shopping list with links to the cheapest or closest hardware stores.
Voice interaction: Let users talk to the bot hands-free while working.
Renovation guidance: Not just repairs—planning upgrades, measuring materials, and debugging renovation mistakes.
Safety overlays: Highlight dangerous components directly on the uploaded image.