Glow Up

Inspiration

We built Glow Up because most people don't actually know what suits them. Fashion apps give generic advice, everyone gets the same tips. We wanted something that answers a more personal question: what would actually look good on me?

How We Built It

Glow Up runs on a multi-agent AI pipeline instead of one model doing everything. The first agent studies the user's profile and builds initial styling picks. The second agent checks those picks against weather, season, and occasion. The third agent validates everything against the user's city and local clothing norms. The last agent turns all of that into detailed, image-ready outfit descriptions.

Those descriptions then get fed into the YouCam API, which is what actually brings the recommendations to life. Instead of just telling the user "this color and cut would suit you," Glow Up shows it, rendering the outfit visually on the user through YouCam. That's the piece that turns this from a styling chatbot into something people can actually see themselves in.

What We Learned

Personalization beats generic advice, by a lot. We talked to 20 women early on and every single one wanted this. That told us we were onto something real. We also learned that a good outfit recommendation isn't just about looks, it has to factor in weather, occasion, location, comfort, and what the person actually likes wearing.

Challenges

The hardest part was making the recommendations feel personal instead of generic. Splitting the reasoning across multiple agents, each with its own job and validation step, is how we solved that. The other challenge was turning styling logic into descriptions detailed enough to actually generate a realistic visual, not just a vague "wear blue" suggestion.

At the end of the day, Glow Up is about making getting dressed less of a guessing game.

Built With

• youcam