Celal Berkay Altunel

AI Wardrobe · One Day Hackathon MVP

Giysem

Giysem turns clothing photos into a digital wardrobe, ranks combinations using color harmony, creates personal virtual try-ons and records outfits on a calendar. The working product was built and submitted between 08:00 and 19:00 in one day.

  • GPT-4o Vision
  • Wiro AI
  • Airtable
  • Next.js
Giysem project visual

01 · The problem

Why it exists

People own clothes but still struggle to see useful combinations across a real wardrobe. A helpful product needs to understand garments, rank combinations deterministically and show the result on the person without making the workflow feel technical.

02 · Ownership

What I built

I built the product concept, visual identity, interface, garment workflow, color analysis, recommendation logic and virtual try-on experience during the hackathon.

03 · System

How it works

Vision extracts garment attributes and colors, deterministic harmony logic ranks compatible combinations and a virtual try-on layer visualizes the outfit. The wardrobe and calendar make recommendations reusable after the first session.

04 · Evidence

What is proven

  • Working MVP built in one day
  • Vision based garment color extraction
  • Deterministic harmony ranking
  • Personal virtual try-on
  • Wardrobe and outfit calendar