BOA / AI BEAUTY SERVICE

From a solo prototype
to a shared AI beauty experience.

I built Face, then brought its direction into BOA.
My focus: authentication and AI Hair Studio.

A four-person capstone connecting AI-assisted styling with member, guest, and kiosk experiences.

Role

Authentication & AI Hair Studio

Period

Sep?Oct 2026

Team

4 members ? Capstone

A personal experiment became
the foundation for a team project.

Face began as my solo AI skin-analysis and beauty-service prototype. The direction was selected by team vote and recommended by our professor, then developed into BOA in a four-person capstone.

SOLO PROTOTYPE

Face

FastAPI ? Next.js ? PostgreSQL
MediaPipe ? OpenCV ? Gemini API
Docker ? AWS EC2

→
TEAM CAPSTONE

BOA

Four-person team ? FastAPI monorepo
My modules: authentication
and AI Hair Studio

My contributionBuilt the original Face prototype and owned BOA?s member, guest, and kiosk authentication module plus AI hairstyle synthesis, transfer, and browser-based dyeing.

One service.
Different ways to enter.

Implemented the full authentication module for member, guest, and kiosk use, with Google and email login and support for four interface languages.

Member access

Google and email login for registered users, implemented within the shared FastAPI application.

Google loginEmail login

Guest & kiosk access

Authentication flows covering guest and kiosk usage alongside the member experience.

MemberGuestKiosk
3Access contexts
2Member login methods
4 languagesKorean ? English ? Chinese ? Japanese

Generate a new hairstyle.
Bring it back to the original face.

Combined Gemini hairstyle synthesis with MediaPipe-based hairstyle transfer onto the original face. Added 40-color dyeing in the browser for further exploration.

Original face

The source image for
the hairstyle experience.

→
Generate & transfer

Gemini synthesis
MediaPipe-based transfer

→
Explore hair color

40 dye colors
applied in the browser

Conceptual workflow illustrations, not screenshots or generated output. Swatches illustrate a selection of colors.

AI features inside
a shared product.

Personal-data context

The project combines account access and face-image features. A technical write-up will describe the implemented image lifecycle and access boundaries once verified against the code.

Project evidence

Face screenshots are available; its server is offline. Screenshots and implementation links will be added to document the prototype and BOA contributions.

This page describes implemented feature scope. It makes no claim about clinical validity or measured skin-analysis accuracy.