StyleAI

StyleAI Studio — AI Beauty & Virtual Try-On Suite

Inspiration

fashion brands spend thousands on physical test photoshoots before knowing if garments suit their models. We built StyleAI Studio to solve this: a unified platform where a single selfie unlocks personalized beauty color science, hair/eye styling, and multi-category virtual try-on for both shoppers and brands.

How We Built It

Backend & API: Built with Python and FastAPI, orchestrating Perfect Corp / YouCam Cloud S2S APIs for neural vision tasks. Cloud & Storage: Asynchronous polling queue with Backblaze B2 object storage for high-speed image processing.

Challenges We Faced

Managing varying AI endpoint latencies (2s to 12s) without freezing the web interface was our biggest hurdle. We solved this by building an asynchronous background task queue with automated image normalization and lightweight polling endpoints for real-time progress updates.

Built With

• fastapi
• youcam