Echo

A Tinder-style mobile marketplace that makes second-hand shopping addictive while reducing fashion waste through AI-powered personalization and sustainability scoring.

Comment

Echo: Sustainable Second-Hand Marketplace

Inspiration

Shopping is hard—second-hand shopping is even harder. But what if it was as easy as scrolling through TikTok?

The fast fashion industry is one of the largest contributors to environmental pollution, producing 10% of global carbon emissions and generating massive textile waste. While 97% of what is used to make clothing can be recycled, only about 13% is, with the rest ending up in landfills. Meanwhile, consumers increasingly want to reduce their environmental impact, with 62% of Gen Z preferring to buy from sustainable brands.

We developed Echo, a mobile marketplace that revolutionizes second-hand shopping by making sustainable fashion accessible, personalized, and engaging. Our platform aims to extend the lifecycle of clothing and other items, reducing waste while creating economic opportunities for sellers.

What does our solution do?

Echo is a mobile marketplace specifically designed for sustainable second-hand shopping, featuring:

1. AI-Agent Experience

  • Personalized AI Stylist: Chat with "Echo," an AI Agent integrated with product database and user preferences that helps users discover sustainable fashion matching their style preferences, where you can add your "liked" product to ask any question.
  • Smart Search & Recommendations: Uses RAG & ML algorithms to match users with relevant products based on their preferences, browsing history, and sustainability priorities. (beta)
  • Product Authentication: Computer Vision powered analysis of product images to verify authenticity and condition. (beta)

2. Intuitive Tinder-Style Discovery

  • Swipe Interface: Familiar swipe mechanism (right for like, left for pass) makes browsing products engaging.
  • Sustainability Score: Each product displays a sustainability score (0-100) based on materials, condition, and eco-friendliness.
  • Detailed Metrics: Shows specific sustainability badges (Organic, Recycled, Fair Trade, etc.) to educate users.

3. Comprehensive Product Information

  • Condition Verification: Detailed classification of item condition (New, Like New, Good, Fair).
  • Material Tracking: Information about fabric composition and environmental impact.
  • Size Measurements: Accurate size details reduce return rates and associated carbon emissions.

4. User-Friendly Features

  • Messaging System: Direct communication between buyers and sellers.
  • Secure Payments: Integrated secure payment processing with buyer protection.
  • Community Features: Sustainability impact dashboard showing user's environmental savings.

How we built it

Frontend

  • React Native with Expo: Provides cross-platform support (iOS/Android) with a single codebase.
  • TypeScript: Ensures type safety and improves code quality.
  • Swiper Component: Custom implementation for Tinder-style product cards.
  • Animated API: Used for fluid animations and transitions throughout the app.
  • Context API: Manages global state for products, authentication, and theme.

Backend

  • Node.js & Express: Provides a robust API server infrastructure.
  • Supabase (PostgreSQL): Main database for storing user data, products, and transactions.
  • JWT Authentication: Secures user sessions and API endpoints.
  • Zod Validation: Ensures data integrity throughout the application.

AI/ML Integration

  • Product Recommendation Engine: Analyzes user behaviour and preferences based on GCP analytics and Multimodal LLM to suggest relevant items.
  • Sustainability Scoring: CV + Custom weighted algorithm that calculates environmental impact scores for each product.
  • AI Stylist: Natural language processing (OpenAI API) for the AI chat assistant.

Data Management

  • Data Import Scripts: Custom tools to process and import sustainable fashion product data.
  • Image Processing: Optimizes product images for both quality and performance.
  • Supabase Storage: Efficiently manages and serves product images with CDN support.

Challenges we faced

Technical Challenges

  1. Cross-Platform Performance: Optimizing the swipe interface to work smoothly on both iOS and Android required significant testing and refinement.

  2. Image Management: Handling multiple high-quality product images while maintaining app performance was challenging. We implemented an image optimization pipeline and used Supabase's CDN to improve loading times.

  3. AI Integration Without Dependencies: Creating an AI chat feature that works without requiring constant internet connectivity or expensive API calls. We developed a hybrid approach with pre-trained models and local processing for common queries.

  4. Database Schema Design: Designing a flexible database schema that could accommodate various product types while maintaining query performance required careful planning and testing.

UI/UX Challenges

  1. Sustainability Visualization: Finding intuitive ways to represent sustainability metrics without overwhelming users required several design iterations.

  2. Swipe Interface Fatigue: We had to balance the engaging swipe mechanism with traditional browsing options to prevent user fatigue during longer sessions.

Accomplishments we're proud of

  1. Seamless Swipe Experience: Our custom implementation of the swipe interface provides fluid, intuitive product discovery that users immediately understand.

  2. AI Agent Conversation Flow: We created an AI Agent integrated with the product database and user preference that genuinely assists users in finding sustainable items matching their style.

  3. Sustainability Scoring Algorithm: Developed a comprehensive scoring system that evaluates multiple factors to provide meaningful sustainability ratings.

  4. Offline Functionality: Implemented caching strategies that allow users to browse previously viewed items even without internet connectivity.

  5. Performance Optimization: Achieved sub-500ms load times for the main product feed even with high-resolution images.

What we learned

  1. Sustainable E-commerce Complexities: Researching the sustainable fashion market revealed complexities in verification, measurement, and consumer education that informed our design decisions.

  2. React Native Animation Techniques: Developed expertise in performance-optimized animations that enhance user experience without draining device resources.

  3. Effective Data Normalization: Learned how to efficiently structure data relationships for complex e-commerce scenarios.

  4. User Behavior Patterns: Through testing, we discovered unexpected patterns in how users interact with sustainability information that helped us refine our interface.

What's next?

Short-term Improvements (1-3 months)

  1. Advanced AI Features: Enhance the AI stylist with more personalized recommendations based on user interaction history.

  2. AR Try-On: Implement augmented reality features allowing users to virtually try on clothing items.

  3. Carbon Footprint Calculator: Add detailed emissions savings calculations comparing second-hand purchases to new items.

Scaling Plan (1-2 years)

  1. Seller Verification System: Implement a robust verification process for sellers to increase trust in the platform.

  2. Community Features: Develop groups and forums for sustainable fashion enthusiasts to share tips and recommendations.

  3. Partnerships: Establish relationships with sustainable brands for verified second-hand sections.

Long-term Vision (3-5 years)

  1. Global Expansion: Scale to international markets with localized sustainability metrics.

  2. Circular Economy Hub: Evolve from marketplace to complete circular economy platform with repair services, upcycling workshops, and educational content.

  3. AI-Driven Complete Outfits: AI model scanning over a 100k+ product database to recommend entire outfits, removing worries about sizing, condition, and coordination.

  4. 3D Spatial Avatar: Users can visualize outfits on a personalized 3D avatar with exact measurements to ensure perfect fit and styling before purchase.

Our technology is designed to scale efficiently with microservices architecture, allowing independent scaling of different components. The recommendation engine is built to continuously improve as our user base and product database grow, making the platform more effective over time.

Built With

Share this project:

Updates