Inspiration

What it does## Inspiration

SkinProof Commerce OS was built for the Perfect Corp x Startup World Cup challenge around one idea: a skin scan should not end as a static score. In beauty retail, shoppers need confidence about what fits their skin, and brands need a trustworthy reason to continue the relationship after the first recommendation.

What it does

SkinProof turns a Perfect Corp Skin Analysis result into a shopper profile, cosmetic concern explanation, regimen rationale, and retail retention cockpit. A shopper can upload a consented face image, see source-labeled concern scores and visual evidence, then understand why a regimen is suggested. The brand dashboard turns the latest scan signal into a local follow-up action and advisor review path.

How we built it

The app is a Next.js web experience with a server-side Perfect Corp Skin Analysis API path. The scan route uploads the image, creates a Perfect Corp task, polls for the completed result, stores proof metadata locally, and maps the returned concerns into profile, regimen, and dashboard views.

The demo separates what is real from what is demo-labeled. The Perfect Corp scan proof, concern scores, profile result, and visual evidence are real. Sample cohort KPIs, sample history, and local follow-up actions are clearly labeled and do not claim live CRM automation.

Challenges

The hardest part was keeping the product honest while still making it compelling. Skin analysis can easily drift into medical language or fake certainty, so SkinProof stays cosmetic-only and avoids diagnosis or treatment claims. Another challenge was making the API result central to the product story instead of using it as a decorative backend call.

What we learned

A strong retail AI product is not just about generating a result. The real value comes from connecting evidence, explanation, trust, and the next action. Perfect Corp provides the verified skin signal; SkinProof turns that signal into shopper confidence and a reason for the brand to follow up.

What's next

Next steps would include multi-scan comparison, advisor review, deeper catalog rules, and real CRM or commerce integrations. The contest version focuses on the core loop: scan -> profile -> regimen rationale -> brand retention action.

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for SkinProof Commerce OS

Built With

  • eslint
  • local
  • next.js
  • perfect-corp-skin-analysis-api
  • proof
  • react
  • server-side
  • tailwind-css
  • typescript
  • vitest
  • youcam-api
Share this project:

Updates