Inspiration

We were inspired by a problem that feels small at first, but affects people almost every day: even when we own clothes, it is still hard to decide what to wear.

A lot of people have wardrobes full of pieces they forget about, outfits they never fully plan, and uncertainty around whether an item will actually look good before wearing it. Shopping platforms have become very advanced, but the experience of managing and styling the clothes we already own still feels fragmented. Wardrobe tracking, outfit planning, virtual try-on, and personalized styling are usually separated across different apps, if they exist at all.

We wanted to build something that brings those pieces together into one experience. Instead of encouraging users to keep buying more, we wanted to help them make better use of what they already have. That idea became Bove Closet: an AI-powered digital wardrobe that helps users upload their clothes, virtually try them on, plan outfits, and organize their wardrobe in a smarter and more visual way.

At its core, Bove Closet is about making fashion more practical, personalized, and sustainable.

What it does

Bove Closet is a mobile-first AI wardrobe assistant that helps users digitize, organize, style, and plan outfits from their existing closet.

Users can:

  • Upload clothing items into a digital wardrobe
  • Automatically remove the background from garment photos
  • Categorize items into wardrobe sections such as shirts, dresses, shorts, and pants
  • Save pieces either to their closet or wishlist
  • Upload a full-body photo to use as their try-on base
  • Use AI virtual try-on to see how a garment looks on themselves before wearing it
  • Save generated try-on looks into a history view
  • Like or save favorite try-on results
  • Plan outfits using a calendar-based wardrobe planner
  • Change scheduled outfits for different days
  • Get weather-aware and wardrobe-aware outfit suggestions
  • Explore styling guidance through an AI closet assistant
  • Build travel packing and outfit plans through the trip planner
  • Customize their in-app identity and profile experience

What makes Bove Closet different is that it is not just a closet archive or just a try-on tool. It connects the full loop:

upload clothes -> visualize them -> plan outfits -> save decisions -> reuse your wardrobe better

How we built it

We built Bove Closet as a full-stack application with a mobile frontend and a backend responsible for image processing, storage, metadata, and AI try-on orchestration.

Frontend

The frontend was built with:

  • Expo / React Native
  • TypeScript
  • A custom pixel-inspired visual design system tailored to the app’s fashion/game-like identity

We designed multiple core screens and flows, including:

  • Home
  • Wardrobe / Wishlist
  • Add Item
  • Virtual Try-On
  • Try-On History
  • Calendar Planner
  • Discover / AI stylist
  • Trip Planner
  • Account / Profile

State is managed on the client to support responsive interactions like selecting outfits, scheduling items, switching between screens, and caching user-related wardrobe data.

Backend

The backend was built with:

  • Node.js
  • TypeScript
  • A lightweight server architecture with structured API endpoints

The backend handles:

  • Avatar photo uploads
  • Garment uploads
  • Background removal and garment validation
  • Try-on request handling
  • Try-on history persistence
  • Metadata updates and deletion
  • Image storage access and signed retrieval

Storage and data

We used:

  • Cloudflare R2 for storing uploaded images and generated assets
  • SQLite for metadata such as garments, avatars, and try-on results

This gave us a simple but reliable way to separate image storage from structured app data.

AI / image pipeline

The most technically significant part of the product was the image pipeline.

For garments:

  • Users upload a clothing image
  • The backend removes the background locally
  • The resulting cutout is validated to make sure it is usable
  • Both the original and processed version are stored

Built With

Share this project:

Updates

Submission history