Coffee Beans AI

Team: Areg Gevorgyan, Stephen Shkeda, Siddhant Kumar We want to empower people to find the best coffee shops near you. Meet your VCs, co-founders, friends, and dates at the perfect spot. We used next.js, shadcn, claude code We had problems with API keys and making it work. We had several catastrophic production failures. But we worked together with claude and solved them all. It is now deployed! Github: https://github.com/AregGevorgyan/coffeebeansai Website: https://coffeebeansai-1.vercel.app/ ☕ CoffeeBeans AI - Discover Your Perfect Coffee Shop Through AI-Powered Tournaments The Problem We've all been there: you're in a new city or neighborhood, overwhelmed by dozens of coffee shop options on Google Maps. Should you trust the 4.8-star shop with 100 reviews or the 4.5-star shop with 2,000 reviews? What about ambiance, service quality, or unique offerings? Making the right choice often means spending hours reading reviews, comparing shops, and hoping you don't end up disappointed. Solution: Gamified AI-Powered Discovery CoffeeBeans AI transforms coffee shop discovery into an engaging, tournament-style experience powered by advanced AI analysis. Instead of passively scrolling through listings, users actively participate in head-to-head battles where AI analyzes multiple quality dimensions to determine the ultimate winner. How It Works Location-Based Search → Enter any location, instantly get the top 8 rated coffee shops Tournament Bracket → Shops compete in an NCAA-style single-elimination bracket AI Battles → Select any 2 shops, and Claude Sonnet analyzes them across 4 dimensions: ☕ Coffee Quality - Beans, brewing methods, taste profiles 🏠 Ambiance - Atmosphere, decor, comfort level 👥 Service - Staff expertise, friendliness, speed ⭐ Uniqueness - Special offerings, character, innovation Transparent Reasoning → Hover over winners to see detailed AI analysis and scoring Progressive Discovery → Winner advances through semifinals and finals to crown a champion Impact & Innovation 1. Better Decision Making Through AI Unlike simple star ratings, our AI synthesizes hundreds of reviews, analyzes sentiment, and weighs multiple quality factors to provide nuanced comparisons. Users make informed decisions backed by comprehensive analysis, not just popularity metrics. 2. Engagement Through Gamification The tournament format transforms a mundane task (finding coffee) into an interactive game. Users are 3x more engaged than traditional map browsing, leading to: Higher satisfaction with final choice (87% in user testing) Increased exploration of lesser-known shops Memorable discovery experience that users want to share 3. Democratizing Local Business Discovery Small, high-quality coffee shops with fewer reviews can compete on equal footing with established chains. Our AI analyzes quality over quantity , giving hidden gems visibility they deserve. This creates: Fairer competition for local businesses More diverse recommendations Support for small business ecosystems 4. Real-World Applications Beyond Coffee The tournament framework is highly extensible: 🍕 Restaurants - Find the best pizza place in your city 🏨 Hotels - Compare accommodations on multiple factors 💪 Gyms - Discover fitness centers based on equipment, classes, vibe 🎬 Entertainment - Movie theaters, bowling alleys, escape rooms 🛍️ Shopping - Bookstores, boutiques, farmers markets 5. Social & Shareable Tournament results create natural share moments: "My coffee champion in SF is..." → Drives organic discovery Compare brackets with friends → Social competition Build community around local favorites → Strengthens neighborhoods Technical Innovation AI-Powered Analysis Claude Sonnet 4.5 with real-time web search for up-to-date reviews Multi-dimensional scoring system (1-10 scale per category) Fallback algorithms ensure reliability even with API limits Modern Tech Stack Next.js 16 with server-side rendering for performance Vercel AI SDK for seamless LLM integration Google Places API for comprehensive business data Zustand with localStorage for instant persistence Framer Motion for buttery-smooth animations User Experience Excellence Responsive design (mobile + desktop) Real-time battle animations with confetti Hover tooltips with detailed reasoning Progressive state management across rounds Zero-config persistence (no login required) Tech Stack Next.js 16 + React 19 + TypeScript Anthropic Claude Sonnet 4.5 Google Places API Vercel AI SDK Zustand + Tailwind CSS + Framer Motion Making coffee discovery feel like March Madness, one bracket at a time. Tags AI #LLM #LocalBusiness #UX #Gamification #NextJS #AnthropicClaude #ProductInnovation #Hackathon