PayNGo

AI travel companion that discovers authentic spots, compares multi-modal transit fares across Rapido, Uber & Metro, and completes 1-click bookings with PayPal Orders API v2.

Comment

Inspiration

Traveling to unfamiliar cities is exhilarating, but navigating local experiences is fragmented and exhausting. A tourist or newcomer in India often juggles 5+ different apps: Google Maps for directions, TripAdvisor for places, Rapido, Uber, or Ola to compare fare surges, and clunky local booking portals with confusing payment gateways.

We asked ourselves: What if an intelligent AI companion could understand your natural language desires, find authentic local spots, calculate the cheapest way to get there across all transit modes, and let you complete the payment safely in a single click using PayPal?

That vision became PayNGo — from "I don't know where to go" to "Here is the best spot + the cheapest way to get there + paid via PayPal" in under 30 seconds.

What it does

PayNGo is an autonomous full-stack AI travel companion and instant checkout engine:

🎯 Smart Newcomer Discovery Radar: Explores curated, verified heritage monuments, cultural hotspots, hotels, and authentic dining across all 28 Indian States and 8 Union Territories. 🚖 Real-Time Multi-Modal Fare Comparison: Computes distance, duration, carbon footprint, and live fare estimates across Rapido Bike (fastest), Ola Auto (best value), Uber Cab, City Metro (eco-friendly), and Walking. 💳 Seamless PayPal Orders API v2 Integration: Server-side Create Order API generates 256-bit encrypted checkout authorizations. Server-side Capture Order API finalizes transactions and verifies capture status. Generates verified digital boarding passes with scan-ready QR codes and printable PDF receipts. 🤖 Agentic AI Travel Assistant ("Travelo"): Conversational assistant powered by Google Gemini that understands complex constraints ("Find me pure-veg dinner under ₹500 in Hyderabad and book transport") and provides one-click PayPal booking triggers. 🧳 Persistent Travel Dashboard: Tracks all confirmed trips, boarding passes, and receipts powered by Supabase PostgreSQL & Auth with offline demo resilience.

How we built it

Core Framework: Built with Next.js 14 (App Router) and TypeScript for fast server-side rendering and type-safe API routes. Payments: Deep integration of the PayPal Orders API v2 using @paypal/react-paypal-js for smart checkout buttons and Node.js backend routes for cryptographic authorization and server-side capture. Database & Auth: Supabase PostgreSQL for real-time relational storage of bookings, profiles, and itineraries, backed by optimistic client caching. Routing & Maps: Interactive geolocation radar using Leaflet and real-time road path geometry powered by the OSRM (Open Source Routing Machine) Engine. AI Agent: Google Gemini 2.0/1.5 Flash integrated with custom travel planning tools and Pan-India spatial datasets. UI/UX: Custom glassmorphic design system crafted with Tailwind CSS, Lucide React, and Framer Motion animations.

Challenges we ran into

Multi-Modal Fare Normalization: Comparing transit prices between disparate transport modes (e.g. per-km metro tiers vs. dynamic cab surge pricing vs. bike taxis) required building a custom fareEngine algorithm accounting for base fares, traffic multipliers, and city-specific tariffs. Strict Payment Security vs. Agentic Autonomy: While AI can propose bookings, automated commerce must never allow silent charges. We architected a strict security boundary where the AI generates order drafts, but execution requires server-side validation and explicit biometric/PayPal user approval. SSR Compatibility with Map Canvases: Leaflet relies heavily on the browser window object. We resolved server-side hydration mismatches using dynamic SSR-exempt module imports (next/dynamic with ssr: false). Judge-Resilient Sandbox Mode: Ensuring that hackathon judges evaluating the app without pre-configured sandbox credentials can still test the complete end-to-end payment and ticketing flow without errors.

Accomplishments that we're proud of

100% Pan-India Coverage: Indexed and tested 68 prominent destinations and transport hubs across all 28 States & 8 Union Territories. Flawless PayPal v2 Lifecycle: Completed a secure 3-step payment flow: Server Order Creation ➔ Client Approval ➔ Server Capture Verification ➔ Dynamic QR Digital Voucher. Speed & Usability: Shipped 16 clean routes with sub-second page loads and zero TypeScript or lint errors. Universal Evaluation Resilience: Built a smart fallback architecture that gracefully handles live Supabase and PayPal credentials while maintaining complete functionality in demo mode. What we learned

What we learned

Deep mastery of the PayPal Orders API v2 security requirements, webhooks, and idempotent capture handling. How to architect robust AI tool-calling workflows that separate natural language intent from actual financial transaction execution. Managing asynchronous map rendering and spatial geometry in modern Next.js React Server Components.

What's next for PayNGo

IRCTC & Airline Direct Booking Integrations: Extending the checkout engine from local city transit to real-time PNR generation for Indian Railways (Vande Bharat Express) and domestic flights. Split Bills with PayPal: Allowing travel groups to automatically split hotel and cab receipts among friends directly via PayPal accounts. Offline PWA Passbook: Storing encrypted boarding pass QR codes in Apple Wallet and Google Wallet for internet-free verification at venue gates.

Built With

Share this project:

Updates

Submission history