Butter Skin: AI Skincare Coach
Analyze your skin with AI, track progress, and get a personalized skincare plan.
Problem Overview
People struggle to find skincare routines that actually work for their unique skin types. Dermatologist visits are expensive, and tracking the efficacy of various products over time manually is tedious and often inaccurate.
Solution & Architecture
Butter Skin utilizes on-device AI and computer vision to analyze skin conditions directly from the user's camera. It generates personalized routines, tracks historical progress via a visual diary, and recommends scientifically-backed products tailored to individual needs.
Key Contributions
- Architected the entire Flutter mobile application from the ground up.
- Integrated TensorFlow Lite for real-time, on-device skin analysis.
- Built a robust state management system using Riverpod for complex user flows.
- Designed and implemented the custom calendar and progress tracking UI.
- Set up Firebase Auth, Firestore, and Cloud Functions for backend synchronization.
Technology Stack
Frontend: Flutter, Dart, Riverpod, Freezed
Backend: Firebase Auth, Cloud Firestore, Cloud Functions
Other: TensorFlow Lite, RevenueCat, GitHub Actions CI/CD
Key Metrics
- < 2s — AI Analysis Time
- 99.9% — Crash-free sessions
- 4.8★ — User Rating