Inspiration
What it does## Inspiration
Kintsugi is inspired by the Japanese art of repairing broken pottery with gold—celebrating an object’s history through its repairs. We wanted to bring that spirit to clothing and cherished objects, making repair feel creative, personal, and approachable.
What it does
Upload a photo of a damaged garment, review the detected damage, and choose a repair design. Kintsugi creates printable repair patterns and step-by-step guidance to help you mend it yourself. For those who need a skilled hand, it also supports finding a maker and requesting a paid repair. Decorative ceramics have their own guided repair flow.
How we built it
We built the studio with React and a FastAPI backend, using private photo storage and background workers for image processing. OpenAI vision assesses damage and suggests repair approaches, while segmentation traces the repair area. The app turns saved designs into downloadable PDF and SVG guides and supports maker quotes, repair progress, shipping, and payout reconciliation.
Challenges we ran into
Turning a photograph into a useful physical repair plan required careful handling of uncertainty. Prints, seams, and shadows can resemble damage, and photographs alone cannot reliably establish physical size. We added editable boundaries, optional measurements, clearly labeled suggested template sizes, and print calibration marks.
Accomplishments that we're proud of
We connected damage assessment, personalized repair designs, printable guides, and illustrated instructions into one repair journey. Users can save their progress or carry their design and completed steps into a maker request when they need help.
What we learned
A compelling repair preview is only part of the solution. Useful guidance also needs clear measurements, practical steps, and honest communication about what AI can infer. Designing for beginners taught us to make every stage understandable and easy to revisit.
What's next for Kintsugi
We want to validate the guides with experienced makers, improve damage detection across more fabrics, expand repair designs, and refine sizing and calibration. We also plan to grow the maker network and use feedback from completed repairs to improve the experience.
Log in or sign up for Devpost to join the conversation.