Search for what you want to buy
Product recommendations
Cart
AR View

Fynd: Find Fast, Fynd Smart

Fynd empowers users to make faster, smarter decisions by combining AI-driven product aggregation with an intuitive interface, reducing decision fatigue and endless scrolling. Here’s a breakdown of our project:

Inspiration

Modern shoppers face overwhelming choices across e-commerce platforms, leading to decision paralysis. We wanted to streamline this process by leveraging AI to curate personalized recommendations while introducing a tactile, engaging interaction model inspired by dating apps. Integrating AR via Snap Spectacles adds a layer of immersive exploration.

What it does

AI-Powered Curation: Agentic AIs crawl multiple e-commerce platforms to gather product data.
Swipe-to-Decide: Users swipe right to "like" products or left to skip, with preferences refining real-time recommendations.
AR Integration: Snap Spectacles enable hands-free browsing, previewing products in augmented reality.
Recommendation Engine: Aggregates liked items to suggest tailored options, minimizing scrolling.

How we built it

Frontend: React-based web app with a card-based UI for swiping (button-based fallback due to hardware constraints).
AI/Backend: Fetch.ai agents for data crawling and recommendation logic, Dain AI + Butterfly for AR visualization.
Hardware: Snap Spectacles for AR previews (basic integration due to time limitations).
APIs: Custom connectors for Shopify, Amazon, and Etsy to aggregate product data.

Challenges we ran into

Snap Spectacles Integration: First-time developers struggled with AR hardware setup and gesture recognition, leading to a button-based swipe fallback.
Multi-Platform Sync: Ensuring real-time consistency across three frontend interfaces (web, mobile, AR) was complex.
Time Constraints: Balancing feature scope with hackathon deadlines forced prioritization of core functionalities.

Accomplishments that we're proud of

Built a functional MVP with three interconnected frontends in 48 hours.
Successfully integrated Fetch.ai agents with live e-commerce data streams.
Created a prototype AR experience despite hardware learning curves.
Achieved seamless handoff between AI curation and user interaction.

What we learned

Hardware Limitations: Developing for AR glasses requires specialized SDK expertise.
Agentic AI Design: Training AI to balance user preferences with diverse product catalogs is nuanced.
Team Dynamics: Rapid prototyping demands clear role delegation and iterative testing.

What's next for Fynd

Expand Use Cases: Apply the framework to restaurants, travel, and recipes.
Enhanced AR: Implement gesture-based swiping with improved Spectacles integration.
AI Optimization: Refine recommendation algorithms using reinforcement learning.
Social Features: Shareable "collections" and collaborative decision-making.
Cross-Platform Support: iOS/Android apps and broader e-commerce API coverage.

Fynd reimagines decision-making as a dynamic, interactive experience-bridging AI efficiency with human intuition.