AI trip planner
Destination, dates, travellers, ₹ budget, interests and pace → day-wise timeline with real places, costs, stays, tips and a budget donut. Server-side validation keeps totals consistent.
Open sample itinerary →
Working demo · AI is live
Your AI travel app,
Your AI travel app,
first version in 20–25 days.
Before you read our proposal, try the product. This demo plans a real day-by-day trip with Amazon Nova Pro on AWS, answers travel questions in a concierge chat, takes you through a booking checkout and shows every booking in an admin dashboard.
- Real AI, not mock-ups: each itinerary is generated live (4–10 s)
- Built for Indian travellers: ₹ budgets, IRCTC/UPI-ready flows, GST in checkout
- Honest labels: flights, hotel rates and payments are simulated until your partner APIs are connected
↑ This is the live app. Tap See a sample or Plan with AI.
4–10 sAI itinerary generation
0API keys in the browser (IAM role)
3surfaces: app, concierge, admin
Desktop + mobiletested in Chromium & WebKit
What's working in this demo
Your FRD and Claude-made front-end design will replace our placeholder brand "Wanderloom". The architecture underneath is what we'd ship.
AI concierge chat
Context-aware: it knows the trip you just generated. Short, practical answers in rupees. Guardrails for visas & payments.
Chat now →Booking checkout
Stay selection, promo code, GST, UPI / card / EMI choice and a booking confirmation. Payment is simulated (Razorpay/Cashfree in production).
Book the sample →Admin dashboard
GMV, bookings, AI usage & latency, top destinations, a searchable bookings table with status updates, and an AI itinerary review queue.
Open dashboard →Secure by default
AI runs in AWS Lambda with an IAM role (no keys in the browser), behind CloudFront with HTTPS, security headers and an origin check.
App-ready
Installable PWA today. For v1 we'd ship Android + iOS from one React Native codebase, sharing the same API.
How v1 ships in 20–25 days
Indicative plan; we'll finalise it with you after reading the FRD. Weekly demo builds you can click, not status reports.
- Days 1–3
FRD walkthrough & build plan
Go through the FRD and the Claude design, confirm v1 scope and APIs (flights/hotels/payments), set up repo, CI, staging on AWS.
- Days 4–10
Core app + backend
Auth (OTP), search, AI itinerary & concierge, trips, profile. Your design turned into real screens, pixel-checked against the design.
- Days 11–16
Booking, payments, admin
Partner API integration, payment gateway, notifications, admin panel for bookings, users, content and AI review.
- Days 17–21
Testing & hardening
Unit + API + end-to-end tests, device testing (Android/iOS), load test of AI endpoints, bug-bash with you.
- Days 22–25
Release v1
Play Store / App Store submission, production deploy, monitoring and a handover doc.
"Testing should be done properly"
Agreed. This is how we test, and how we tested this demo.
Automated tests from day 1
Unit tests for business logic (pricing, GST, budget checks), API tests for every endpoint, end-to-end tests for search → plan → book.
Real browsers & devices
This demo was tested with Playwright in Chromium and WebKit on desktop and mobile sizes, with zero console errors and every asset loading.
AI output checks
AI replies are validated against a schema, totals re-computed on the server, and a fallback shown if the model is unavailable.
Proposed stack
React Native (Android + iOS)React / Next.js adminNode.js / TypeScript APIPostgreSQLAWS (Lambda, S3, CloudFront)Amazon Bedrock or OpenAIRazorpay / CashfreeFirebase pushPlaywright + Jest
Open to your preferred stack if the FRD already specifies one.