Inspiration
Choosing what to wear is time-consuming and often frustrating. Most fashion apps either overwhelm users or treat outfits as static sets. We wanted to make styling fast, personalized, and intelligent—like having a stylist that actually learns you.
⸻
What it does
StyleSwipe is an AI-powered fashion app where users swipe through outfits while autonomous agents work in the background to find clothing tailored to their style.
Unlike traditional systems, StyleSwipe: • 👕 Understands individual clothing items, not just full outfits • 🤖 Uses AI agents to continuously discover and refine recommendations • 🔄 Dynamically builds outfits by combining pieces that match your preferences
The result is a system that doesn’t just suggest outfits—it actively curates your style in real time.
⸻
How we built it • Frontend: TypeScript/JavaScript swipe-based UI for rapid feedback • Backend: Python server handling user data, recommendation logic, and agent coordination • AI System: • Autonomous agents for item discovery and outfit assembly • Preference learning from swipe behavior • Item-level feature analysis for precise matching
⸻
Challenges we ran into • Modeling fine-grained preferences at the clothing-item level • Designing agents that balance exploration vs. personalization • Ensuring compatibility between independently selected pieces • Keeping recommendations fast and responsive
⸻
Accomplishments that we’re proud of • Built a working agent-driven recommendation system • Achieved item-level styling intelligence, not just outfit matching • Created a seamless loop where the app learns instantly from user input • Delivered a polished, interactive prototype under hackathon constraints
⸻
What we learned • How to design real-world AI agent systems • Why UX is critical for making AI feel useful • The power of granular data (items vs. outfits) in recommendations
⸻
What’s next for StyleSwipe • Closet integration + real wardrobe syncing • Direct shopping from recommendations • Smarter, specialized agents (e.g., seasonal, occasion-based) • Mobile app + social sharing
Built With
- auth
- fastapi
- gemini
- javascript
- postgresql
- python
- typescript
- vision
Log in or sign up for Devpost to join the conversation.