Your AI-Powered Personal Stylist for Snap Spectacles.

SnapDrobe bridges the gap between technology and fashion. By leveraging AR, voice interaction, and the Gemini API, SnapDrobe allows users to build a digital twin of their wardrobe and receive context-aware outfit recommendations—all hands-free.

Fashion isn't everyone's strong suit. Our team loved technology but struggled with the daily "what do I wear?" dilemma. When we got access to the new Snap Spectacles, we saw a chance to turn a personal pain point into a seamless AR experience. SnapDrobe makes fashion effortless, fun,ed and data-driven by capturing style inspiration the moment you see it.

SnapDrobe transforms how you interact with your closet through simple voice commands:

• Instant Wardrobe Capture: Spot a piece you love? Say "Add to wardrobe" to capture it via Spectacles.
• AI-Powered Analysis: Gemini analyzes the image to store structured metadata (color, material, style) in Amazon DynamoDB.
• Smart Suggestions: Ask for an outfit based on intent (e.g., "I'm going to a rooftop party").
• Context Awareness: Recommendations are tailored to your existing collection, current location weather, and the time of day.
• Visual Previews: View an AI-generated mockup of the full head-to-toe look directly in your AR field of view.

SnapDrobe is a full-stack AR application built by a team of three:

• Lens Studio & TypeScript: Designed the spatial UI and interaction layer.
• Spectacles Interaction Kit: Handled voice-activated triggers and user input.

• FetchAI UAgent Framework: Managed communications between the hardware and cloud.
• Gemini API: Used for both visual analysis (image-to-JSON) and reasoning (outfit generation).
• OpenWeather API: Provided real-time environmental context.
• Amazon DynamoDB: A scalable NoSQL database for personal wardrobe storage.

• The AR Learning Curve: As first-time Lens Studio users, mastering the Spectacles Interaction Kit and spatial UI positioning was a steep climb.
• Prompt Engineering for JSON: Ensuring the Gemini API consistently returned structured data for the database required meticulous prompt refinement to avoid parsing errors.
• Orchestrating Latency: Connecting voice input → weather data → wardrobe retrieval → AI image generation created a complex chain. We had to optimize our FetchAI agents to ensure the user wasn't left waiting in an AR environment.
• Spatial UI Design: Learning to build interactive, responsive interfaces for an AR/VR environment was completely new territory, requiring us to think about "depth" rather than just "pixels."

• Successfully built an end-to-end system connecting AR wearables to cloud AI.
• Navigated a production-ready experience in Lens Studio with no prior AR/VR experience.
• Integrated multiple disparate technologies (FetchAI, Gemini, DynamoDB) into a single fluid workflow.

• AR Try-On: Overlay recommended outfits onto the user’s body in real-time.
• Social Wardrobes: Collaborative styling and "closet sharing" with friends.
• Calendar Integration: Proactive outfit suggestions based on your upcoming meetings or trips.
• Style Learning: Personalizing recommendations based on user preferences over time.

• Hardware: Snap Spectacles (AR/VR)
• AI: Gemini API
• Agents: FetchAI (UAgent)
• Database: Amazon DynamoDB
• External APIs: OpenWeather API
• Development: Lens Studio, TypeScript, Python