LUMÉ — AI-Powered Beauty Intelligence Platform
Inspiration
Beauty and skincare shopping today is still surprisingly uncertain. Most users struggle to understand what products actually suit their skin, how makeup will look before purchasing, or how to build a personalized routine without expensive consultations.
We wanted to create an experience that feels futuristic, intelligent, and confidence-boosting — something that combines AI, beauty technology, and immersive interaction into a single platform.
LUMÉ was inspired by the idea of bringing professional-grade beauty consultation to everyone through AI-powered personalization and virtual beauty experiences.
By combining Perfect Corp’s beauty intelligence technology with conversational AI, we built an experience where users can upload a selfie, receive personalized skin insights, explore curated skincare routines, and virtually try on beauty looks in real time.
What it does
LUMÉ is an immersive AI beauty platform that provides:
- AI-powered skin analysis
- Personalized skincare recommendations
- Interactive beauty scoring
- Virtual makeup try-on experiences
- AI beauty advisor chat assistant
- Tailored morning & evening skincare routines
- Beauty concern analysis and glow recommendations
Users simply upload a selfie, and the platform generates a complete beauty profile within seconds.
How we built it
We built LUMÉ using a modern full-stack AI architecture focused on performance, immersive UI/UX, and scalability.
Frontend
- React
- Next.js
- TypeScript
- TailwindCSS
- Framer Motion
Backend
- Secure API routes
- AI processing endpoints
- Environment-based API security
AI + Beauty Technology
- Perfect Corp APIs
- Claude AI integration
- Personalized recommendation logic
- Dynamic beauty analysis workflows
We focused heavily on:
- cinematic user experience
- glassmorphism-inspired design
- smooth transitions
- interactive beauty flows
- mobile responsiveness
- startup-quality visual polish
Challenges we ran into
One of the biggest challenges was designing an experience that felt both technically impressive and emotionally engaging.
We spent significant time refining:
- smooth AI interaction flows
- responsive animations
- personalized recommendation logic
- secure API architecture
- immersive onboarding experience
Another major challenge was balancing realism with speed during skin analysis and virtual try-on experiences while maintaining a polished UX suitable for live demos.
We also had to carefully structure backend API handling to securely integrate AI services and sponsor technologies without exposing sensitive credentials.
Accomplishments that we're proud of
We are especially proud of:
- building a visually polished AI beauty experience
- creating a startup-quality UI in hackathon time constraints
- integrating conversational AI with beauty intelligence
- delivering personalized skincare guidance
- designing an immersive virtual try-on flow
- creating a platform that feels production-ready instead of a simple prototype
What we learned
Through building LUMÉ, we learned:
- how to architect secure AI-powered frontend/backend workflows
- how to create emotionally engaging AI consumer experiences
- how important UX polish is for AI products
- how beauty technology APIs can create highly personalized retail experiences
- how to combine AI recommendations with interactive visual storytelling
We also learned the importance of designing AI products that users can instantly understand and emotionally connect with.
What's next for LUMÉ
Future plans for LUMÉ include:
- real-time AR makeup overlays
- AI-generated beauty transformations
- personalized product marketplace integration
- influencer-inspired beauty looks
- social sharing and creator features
- advanced skincare progress tracking
- beauty commerce integrations
- multilingual AI beauty coaching
Our long-term vision is to build an AI-native beauty companion that transforms how users discover skincare, makeup, and self-confidence.
Built With
- ai
- apis
- corp
- css3
- framer
- html5
- motion
- next.js
- node.js
- perfect
- react
- tailwindcss
- typescript
- vercel


Log in or sign up for Devpost to join the conversation.