An AI-powered hyperlocal marketplace for local creators, makers, and food entrepreneurs.
Built with GPT-4, Stripe Connect, and a modern React/Node.js stack to simplify local selling and eliminate manual coordination.
- 🤖 AI Pickup Scheduler — GPT-4 suggests optimal pickup times based on both seller & buyer availability, weather, and traffic.
- 🛍️ Smart Listings — GPT-4 auto-generates product titles, descriptions, and tags for local goods like crafts, produce, and baked items.
- 📸 Verified Product Imagery — AI-enhanced product image matcher using Pexels/Unsplash APIs with 90%+ confidence scoring.
- 💳 Seamless Payments — Fully integrated with Stripe Connect Express for secure seller payouts and buyer transactions.
- 🧾 Professional Receipts — PDF receipts with QR code confirmation for every order.
- 🔐 Error Handling & Validation — Robust backend validation, real-time logging, and user authentication.
- Frontend: React, TailwindCSS, TanStack Query
- Backend: Node.js, Express.js, Firebase (Auth + Firestore)
- AI: OpenAI GPT-4 (via API), Pexels/Unsplash APIs for image matching
- Payments: Stripe Connect Express (OAuth + payout flow)
- Deployment: Replit (with GitHub-linked CI/CD)
- Sellers save 6.2 hours/week coordinating pickups
- 92% AI pickup suggestions accepted
- 94% order completion success rate
- All featured products now have verified AI-matched images
Use the "Smart Add Product" feature or explore as a buyer to see the full AI flow in action.
Christopher Crilly Pienaah
🔍 Analytics Student, Northeastern University
🎓 Concentration: Applied Machine Intelligence
📧 Contact: ccpienaah@gmail.com
🌐 LinkedIn: linkedin.com/in/crillypienaah
- 📱 Mobile-first UX & accessibility
- 📦 AI inventory learning + demand prediction
- 📍 Launch local pilot with real sellers
Built in 72 hours during the DTTP x World Innovation League Hackathon. Selected as one of 6 finalist teams (out of 25+).
NeighborCart makes selling local... smart, simple, and stress-free.