Inspiration
19% of online beauty purchases get returned because products look wrong on real skin tones. Nobody had fixed this properly.
What it does
Detects your skin undertone from a photo, virtually tries on lipstick and clothing, and tells you honestly if the color works for your complexion. Also checks skincare ingredient conflicts and detects hair type for product matching.
How we built it
Python + Flask backend, HTML/CSS/JS frontend. Perfect Corp APIs for skin analysis and virtual try-on. Anthropic Claude API for personalized verdicts.
Challenges we ran into
Getting the Perfect Corp API authentication and base URL correct. Chaining multiple API calls per feature without blocking the UI.
Accomplishments that we're proud of
Three working features that each solve a real, documented consumer problem — all in one cohesive app built in under a week.
What we learned
How to chain async REST APIs in Flask. The science behind skin undertones and color compatibility. That warm human language beats raw API scores every time.
What's next for ShadeMatch
Embed ShadeMatch as a widget into e-commerce platforms — reducing returns for fashion retailers and increasing cart confidence for skincare brands.
Log in or sign up for Devpost to join the conversation.