Agent Stitch

Imagine Miranda Priestly personally picking your outfit every single day and Monica Gellar decluttering your closet for you! Wow, talk about special treatment! You will never buy things you don't need

Comment

Inspiration

"I have nothing to wear!", "I forgot my favorite red top was in the back of the closet so I bought another one!" As victims to these thoughts every now and then, we decided to get rid of the hassle altogether! Introducing Agent Stitch, your personalized AI stylist that will not only pick chic outfits for you but will also make sure you're making sustainable wardrobe choices!

What it does

  • A digital closet where users upload clothes via camera. It uses multimodal AI to automatically tag items by type, color, season, and formality, eliminating manual data entry or tagging.
  • We used a persona-driven agentic architecture: Miranda (The Stylist): Generates personalized outfit recommendations by matching inspiration looks against the user’s actual inventory. Monica (The Declutter Agent): Promotes sustainability by identifying unworn items and suggesting candidates for donation. Ralph (The Fashion Chatbot): A conversational agent that provides real-time styling advice and can "see" photos to suggest pairings with new potential purchases.

How we built it

  • Mobile Frontend: Built with React Native (Expo) and the mobile lifecycle was managed by handling AsyncStorage for persistence and expo-image-picker for seamless hardware integration.
  • AI Engine: Integrated Google Gemini 2.5 Flash (Vision + Text) via a FastAPI backend.
  • Persistent Memory: SQLite (SQLAlchemy) was used to store wardrobe data and saved outfits, ensuring the AI has a "memory" of the user's closet.
  • Contextual Awareness: By integrating the Open-Meteo API, the system automatically detects the current season to ensure outfit suggestions are weather-appropriate.

Challenges we ran into

  • One of our primary technical hurdles was implementing persistent context memory for our chatbot. Ensuring the agent consistently recalled user profiles, conversation history, and uploaded images required meticulous state management. Additionally, transitioning from web to mobile app development presented a steep learning curve.

Accomplishments that we're proud of

  • One of our proudest achievements is our refined, minimalist UI. Designed to feel high-end and effortless, the interface maintains its elegant aesthetic even in dark mode, making late-night outfit planning feel as beautiful as the clothes themselves.
  • Beyond aesthetics, we empower sustainable fashion through our AI chatbot. Before you buy, simply ask if an item is necessary; the agent scans your entire digital wardrobe to identify similar pieces and instantly calculates how many existing items will pair perfectly with your new find.

What we learned

  • The development of Agent Stitch highlighted the need for a shift to mobile paradigms, the necessity of multimodal context for AI assistants, and the importance of managing hardware-software orchestration for performance.

What's next for Agent Stitch

Future development will focus on collaborative styling features with friends and community, e-commerce integration, automated wear-analysis using an avatar, and migrating to a production-ready cloud infrastructure.

Built With

Share this project:

Updates