Aurafit

Aurafit is an AI personal stylist that creates personalized outfits using real products you can actually buy, based on your appearance, occasion, budget, and style preferences.

Comment

Aurafit

Demo Access

Aurafit is currently hosted on Azure and can be brought online for judging upon request. To reduce infrastructure costs during development, the production deployment may be paused when not actively being evaluated.

A complete walkthrough video and Android APK are provided for evaluation.

Inspiration

Choosing what to wear is a problem almost everyone faces. Whether it's a date, an interview, a wedding, or even a casual outing, people often spend a surprising amount of time trying to decide what looks good on them.

Most fashion apps either show generic recommendations or generate AI outfits that don't actually exist. We wanted to build something more practical — a personal stylist that understands the user and recommends outfits built from real products that can actually be purchased.

That idea became Aurafit.

What it does

Aurafit is an AI personal stylist that helps users decide what to wear.

Users upload a photo, select their occasion, budget, and style preferences, and Aurafit generates multiple personalized outfit recommendations tailored to them.

Unlike traditional fashion AI tools, Aurafit doesn't stop at generating outfit ideas. It finds real products, builds complete looks around them, and provides direct shopping links so users can actually purchase the outfits they like.

Aurafit also includes a wardrobe feature that allows users to upload clothes they already own and receive recommendations built around their existing collection.

How we built it

We built Aurafit as a full-stack AI-powered fashion platform.

The frontend was built using React, Next.js, Tailwind CSS, and React Native for the mobile experience.

The backend uses PostgreSQL, Azure Storage, and Azure OpenAI to handle user data, image analysis, outfit generation, and recommendation workflows.

To discover products, we built a product discovery pipeline using Playwright and Crawl4AI. This allows Aurafit to find relevant products, gather product information, and create outfit recommendations using real items available online.

The system combines user analysis, outfit planning, product discovery, and image generation to create a complete styling experience.

Challenges we ran into

One of the biggest challenges was maintaining the user's identity across generated looks. We wanted recommendations to feel personal and consistent rather than producing completely different-looking people in every outfit.

Another challenge was building a scalable database and storage architecture that could handle user profiles, wardrobes, generated outfits, and recommendation history efficiently.

Product discovery was also difficult. Many shopping websites actively block traditional crawlers and scraping tools, making it challenging to reliably collect product information and images. We had to experiment with different crawling approaches and build a more robust product discovery workflow to maintain recommendation quality.

Finally, balancing generation quality, response speed, and infrastructure cost required significant iteration throughout development.

What we learned

This project taught us far more than just building an AI application.

We gained hands-on experience working with cloud infrastructure, Azure services, storage systems, databases, and production deployment workflows. We learned how real-world applications differ from classroom projects, especially when reliability, scalability, performance, and cost become important considerations.

We also learned the challenges involved in designing systems that can grow beyond a prototype. From managing storage and databases to handling AI workloads and product discovery pipelines, every component needed to be designed with scalability in mind.

Most importantly, we learned that building a useful product is not just about creating a working model. It is about combining multiple systems into an experience that people can actually use and benefit from in the real world.

What's next

Our goal is to make personal styling accessible to everyone.

We plan to improve wardrobe intelligence, expand product coverage, introduce smarter outfit personalization, and continue building Aurafit into an AI stylist that helps users make confident fashion decisions every day.

Built With

Share this project:

Updates