Powered by a dedicated on-device AI chip, all spectral data is processed locally — completing the sense-analyze-act cycle within 50 milliseconds.
Multi-Spectral Sensing,
Deep Insights into Skin Cells
Multi-wavelength detection from 450nm to 850nm accurately identifies chromophores such as melanin, hemoglobin, and collagen — revealing 12 types of skin concerns.
Lightweight Machine Learning,
Precision with Low Power
A lightweight combination of 1D-CNN with attention mechanism and optimized TabNet.
Continuous Learning,
Smarter with Every Use
On-device continuous learning reduces model size by an additional 30%, running faster and more efficiently — freeing up compute power for precision skincare.
Covering global skin types and tones, annotated by professional dermatologists — building a dual moat of technology and data.
