Skincare advice has traditionally meant an expensive appointment or trial and error. We built ClearSkin AI to put a considered reading of your own skin in between those two options.
Understanding your own skin should not require a referral and a three-month wait. ClearSkin AI uses a vision model to give anyone with a smartphone a structured, repeatable reading of their skin — the same six metrics, the same four heatmap categories, every time.
We are not a replacement for a dermatologist and we do not pretend to be. What we can do is close the gap between knowing nothing and booking an appointment, and give you something concrete to bring with you when you do.
Three views of your face assessed for acne, dryness, oiliness and redness, resolved into a comprehensive score in under thirty seconds.
Visual overlays across four categories show exactly where concerns are concentrated rather than only that they exist.
An estimated skin age against your actual age, with a confidence figure, tracked alongside your overall score over time.
Any two scans side by side, with delta indicators showing which metrics moved and in which direction.
AM and PM routines built around what your scan actually found, with targeted product suggestions.
Row-level security on every table, encryption in transit and at rest, compound rate limiting, OWASP Top 10:2025 compliant.
React Native and Expo, in TypeScript, with NativeWind for styling.
A vision-capable model assesses each captured angle and returns structured metrics.
Supabase — Postgres, auth, storage, edge functions and realtime.
App Store and Google Play billing, with RevenueCat for subscription status.
Resend for transactional email and contact form delivery.
Jest with React Native Testing Library across 29 test files.