Agentic AI for OLA,Uber,Rapido

Agentic AI Ride Booking App

Agentic AI Ride Booking App is a full-stack, AI-powered platform designed to simplify ride booking. Instead of manually checking multiple ride-hailing applications for the best price or fastest arrival time, users can simply describe their requirements in natural language. The intelligent AI agent automatically compares live data across Uber, Ola, and Rapido, selects the best option based on the user's preference (such as cheapest or fastest), and books the ride.

Key Features

Conversational AI Interface:
Users can book rides through a chat-based interface powered by Anthropic's Claude AI.

Real-Time Cross-Platform Comparison:
The system simultaneously fetches and compares ride prices, ride types (Bike, Auto, Cab), and estimated arrival times from Uber, Ola, and Rapido.

Smart Rankings and Explanations:
The AI not only books the ride but also explains the reason behind its selection. For example, it may inform the user that an Ola Auto was booked because it was 20% cheaper than Uber for the same route.

Live Ride Tracking:
After booking confirmation, users can track their ride on a map with WebSocket-based real-time driver movement simulation.

Mock Mode / Sandbox Environment:
The application uses realistic mock data, allowing it to function effectively for testing and demonstrations without requiring actual API keys from ride-hailing providers.

Technical Architecture and Technology Stack

Frontend (Web and Mobile)

The web application is built using React and Vite, with Tailwind CSS providing a responsive and modern user interface.

The mobile application is developed using Expo (React Native) and React Native Maps to deliver a native mobile experience.

A shared codebase is maintained through a packages/shared directory, which contains TypeScript types, custom hooks, and utility functions used across both web and mobile platforms.

Backend Server

The backend is developed using Python and FastAPI, providing high performance and efficient API handling.

Anthropic's Claude AI is integrated for intent recognition, tool usage, and automated ride selection.

WebSockets are used for real-time communication and live location streaming through the /ws/{ride_id} endpoint.

RESTful APIs are exposed through /api/* endpoints for ride estimates, chat processing, booking management, and ride history.

Database and Infrastructure

PostgreSQL is used for storing user information, ride history, and application logs.

Redis is used for session management, caching, rate limiting, and real-time event broadcasting.

Supabase Authentication provides secure user login and session management.

How It Works

1. Request
The user enters a request such as "Find me the fastest bike to Indiranagar."

2. Analysis
The FastAPI backend sends the request to Claude AI for intent analysis.

3. Comparison
The AI agent uses internal tools to fetch ride estimates from Uber, Ola, and Rapido integrations.

4. Decision
The system compares available options and identifies the best ride based on the user's preference.

5. Execution and Tracking
The selected ride is booked automatically, and a WebSocket connection is established to stream the driver's real-time location to the frontend map.

Conclusion

This project demonstrates the use of Agentic AI, real-time communication, and full-stack development to create an intelligent ride-booking assistant that automates decision-making and enhances user convenience.