AuraMatch

Home Page
Wardrobe Section for uploading clothes
Outfit Planner and gives suggestions
Smart Shop Recommender redirects to external E-Commerce sites
Example of redirect E-Commerce site

Inspiration

Choosing what to wear every day is a small but constant problem. Many people own several clothes but still struggle to combine them properly, especially for important events like interviews, festivals, or parties. In rural and semi-urban areas, access to fashion advice or stylists is limited. We wanted to build an AI that acts like a personal wardrobe assistant that understands your clothes, your events, and your style, and then suggests the best possible outfit visually.

What it does

AuraMatch is an AI-powered wardrobe planning agent. Users upload photos of their clothes, and the system analyzes each item’s type, color, style, and fabric. Based on the selected event—such as an interview, party, cultural celebration, or casual day—the AI generates and ranks the top outfit combinations.

It also:

Suggests backup outfits if an item is unavailable
Recommends accessories and perfumes
Gives wardrobe improvement tips
Generates a visual preview of the user wearing the selected outfit

This helps users make confident fashion decisions without physically trying every combination.

How we built it

We used Google Gemini as the reasoning engine of the application. Gemini analyzes clothing images, understands style and color harmony, and generates ranked outfit combinations based on the selected event.

The system works in three stages:

Wardrobe scanning: Upload clothing images and extract attributes like type, color, and style.
AI outfit reasoning: Gemini generates and ranks outfit combinations.
Visual output: The app produces a visual representation of the recommended outfit on the user.

The frontend is mobile-first, designed for quick wardrobe uploads and instant suggestions.

Challenges we ran into

One of the main challenges was setting up the Gemini API and cloud billing within the limited hackathon time. We also had to design the logic for ranking outfits based on different events while keeping the system simple and fast enough for a live demo.

Another challenge was generating consistent visual outputs that matched the suggested outfit combinations.

Accomplishments that we're proud of

Building a functional AI wardrobe agent within a short time.
Creating a system that not only suggests outfits but also ranks and explains them.
Generating visual previews to make decisions easier for users.
Designing a mobile-first interface suitable for everyday use.