HackByte 3.0 Top 100 Teams

AgentX Travel India

AI-powered travel assistant with multi-agent system for personalized Indian travel planning

AgentX Travel India

Project Overview

AgentX Travel India is an AI-powered travel assistant application tailored for the Indian market. Built by TechMatrix Solvers team for IIITDMJ HackByte 3.0, it achieved Top 100 Teams recognition.

The application leverages a sophisticated multi-agent system with specialized AI agents for research, accommodation, transportation, activities, dining, and itinerary integration. It features interactive maps, geo-based attraction search, and a conversational chatbot assistant.

Top 100

Teams

4 Agents

AI System

4 Members

TechMatrix Solvers

April 2024

HackByte 3.0

Multi-Agent Architecture

πŸ”

Research Agent

Gathers comprehensive destination information and travel insights

🏨

Accommodation Agent

Finds and recommends suitable hotels and lodging options

πŸš—

Transportation Agent

Plans optimal routes and transportation methods

🎯

Activities Agent

Suggests attractions, activities, and experiences

🍽️

Dining Agent

Recommends restaurants and local cuisine options

πŸ“‹

Itinerary Agent

Integrates all recommendations into cohesive travel plans

Key Features

πŸ—ΊοΈ

Interactive Maps

Visualize travel destinations with integrated mapping using Pydeck for enhanced user experience

πŸ“

Geo-based Attraction Search

Find nearby attractions using MongoDB's geospatial and vector search capabilities

πŸ’¬

Chatbot Assistant

Get answers to travel questions through conversational AI powered by Gemini

πŸ“±

Responsive Design

Works seamlessly across desktop and mobile devices with Streamlit interface

πŸ’Ύ

Downloadable Itineraries

Save travel plans offline for easy access and sharing

πŸ”§

Optional Enhancements

Integration with Tailvy API and MongoDB for improved recommendations

Technology Stack

Frontend & Framework

Streamlit 1.32.2 Python Pydeck 0.8.0

AI & Machine Learning

LangChain Google Generative AI (Gemini) 0.3.2 OpenAI API

Database & APIs

MongoDB Atlas Tailvy API Geopy 2.4.1

Data & Visualization

Pandas 2.1.4 Hugging Face GitHub

Architecture Evolution

v0

Initial Architecture

Basic AI agent integration with simple workflow design

v0.1

Hackathon Implementation

Enhanced architecture with improved agent communication during HackByte 3.0

v1

Advanced Implementation

Advanced version with additional data sources and optimized workflows

v2

Future Vision

Planned future architecture with expanded capabilities and integrations

Team TechMatrix Solvers

Abhay Gupta

Team Leader & ML Developer

Jay Kumar

ML & Front-end Developer

Kripanshu Gupta

MERN Stack Developer

Aditi Soni

Full Stack Developer

HackByte 3.0 Tracks Followed

Gemini API

Primary AI language model for intelligent travel recommendations

MongoDB Atlas API

Database for storing and querying travel data with geospatial capabilities

Cloudflare AI SDK / HuggingFace

Additional AI capabilities and model deployment

GitHub

Version control and collaborative development

Postman Testing

API testing and development workflow optimization

Project Highlights

Multi-Agent System Implementation

Sophisticated AI agent architecture with specialized agents for different aspects of travel planning

Geospatial Intelligence

Advanced location-based search and recommendation system using MongoDB's geospatial queries

Vector Search Integration

Semantic matching of attractions using OpenAI embeddings for personalized recommendations

Interactive Visualization

Real-time map visualization of travel destinations and attractions using Pydeck

Scalable Architecture

Modular design allowing for easy expansion and integration of new features