fitcheck

Inspiration

For many Gen Z users and college students, deciding what to wear is a daily challenge. People want to express themselves and feel confident, but often struggle to style their clothes or decide whether an outfit they saw online would actually suit them.

Social media gives us endless fashion inspiration, but it usually ends at simply viewing or liking a post. We wanted to make that inspiration interactive.

That inspired us to build Fitcheck — a social media app where outfits are the content.

What it does

Fitcheck allows people to:

Post the outfits they are wearing.
Discover looks shared by others through a social feed.
Virtually try on clothes worn by people in their feed.
Save their favorite generated fits.
View and shop the exact pieces tagged in a post.
Create personal clothing collections using shopping links.
Publish new looks and inspire other members of the community.

When someone discovers an outfit they like, they can tap Try, select a saved full-body photo, and use YouCam's virtual try-on technology to see that complete look on themselves.

This creates a continuous social fashion experience:

Discover → Try on → Save or shop → Post → Inspire someone else

How we built it

The mobile application was built with Flutter and Dart. It includes the social feed, profiles, collections, reusable model photos, saved fits, product tagging, likes, comments, and virtual try-on flows.

Our backend uses Node.js and Hono and is deployed on AWS EC2 behind Caddy HTTPS. It manages authentication, posts, uploaded media, collections, saved fits, shopping links, and all AI provider requests.

When a shopping link is added, the backend first attempts to extract structured product information directly from the retailer. If the retailer blocks server requests or serves an incomplete page, the OpenAI API with web search identifies the exact linked product and retrieves its information and product imagery. OpenAI also creates the clean garment asset used in Fitcheck's product cards.

The landing page was built with Next.js and React and deployed on Vercel.

How we use YouCam

YouCam is the core visual engine behind Fitcheck.

YouCam AI Clothes v3 transfers individual garments or complete outfits onto a user's saved full-body photo.
YouCam AI Shoes handles products categorized as footwear.
YouCam Background Replace gives published generated looks a consistent editorial background before they enter the public feed.

The integration runs completely through our backend. It securely uploads the required images, creates the appropriate YouCam task, monitors its progress, downloads the completed result, and stores it behind a stable Fitcheck media URL.

This makes virtual try-on part of the social experience instead of presenting it as a separate tool.

Challenges we faced

One major challenge was transferring an entire outfit from a social post while keeping the result visually consistent. We solved this by sending the composed outfit as a complete reference through YouCam AI Clothes v3 instead of repeatedly editing the image for every tagged piece.

Shopping links were another challenge. Retailers use different page structures, short links, redirects, and anti-bot systems. We combined deterministic product extraction with OpenAI web research so Fitcheck can still identify the exact item without depending on a scraping proxy.

We also had to handle asynchronous YouCam tasks, image validation, temporary result URLs, network failures, and clear user-facing error messages.

Finally, user-uploaded photos had very different framing from our editorial content. We combined YouCam Background Replace with adaptive feed cropping so community posts remain visually consistent.

Accomplishments that we're proud of

We built a working end-to-end social fashion experience rather than a wrapper around a single API call.

A user can discover another person's outfit, virtually try it on, save the result, open its original shopping links, create a new look, and publish it back to the community.

We are also proud that the Flutter app, AWS backend, OpenAI product pipeline, YouCam workflows, and Next.js landing page are all functional and deployed.

What we learned

We learned that virtual try-on becomes much more valuable when it is placed inside an existing human behavior: discovering styles from other people.

We also learned the importance of separating each AI system's responsibility. OpenAI helps understand shopping links and prepare product assets, while YouCam specializes in generating the actual fashion try-on experience.

Building reliable AI features also required more than making an API request. Image validation, asynchronous polling, retries, persistent media, and understandable failure states were all essential parts of creating a usable product.

What's next for Fitcheck

Next, we want to introduce personalized outfit recommendations based on saved fits, creator profiles and follow systems, collaborative collections, wardrobe planning for college events, and additional YouCam accessory experiences.

We also plan to move private media to signed object storage, adopt a scalable managed database, and build direct retailer and creator partnerships.

Our long-term goal is to make Fitcheck the place where people do more than discover fashion—they experience it on themselves.