Team HashHackers

---

🌍 TBOAnalytica: AI Travel Growth & Decision Platform

An AI-powered Travel Operating System that combines:

• Intelligent Itinerary Optimization
• AI Marketing Automation
• Multi-Agent Decision Intelligence
• WhatsApp Campaign Orchestration

This platform transforms TBO from a booking inventory provider into a Growth + Decision Intelligence Engine for Travel Agents.

---

🏗 System Architecture

The system consists of:

Component	Description
itinerary_module	AI-powered itinerary creation, scoring, simulation & conversational reasoning
agent_module	AI-driven campaign intelligence, creative generation & WhatsApp automation
frontend	Web interface for interacting with both modules
PostgreSQL	Centralized relational database

---

🗄 Database Setup

Before running the system, set up PostgreSQL.

Create Database

Open terminal:

psql -U postgres

Create database:

CREATE DATABASE travel_ai;

Exit and import schema:

psql -U postgres -d travel_ai -f Backend/db/schema.sql

This will create all required tables.

ER Diagram

Full Local Setup Guide

You must run three components:

1. Itinerary Backend
2. Agent Backend
3. Frontend

1. Itinerary Backend

cd Backend/itinerary_module
python -m venv myenv
source myenv/bin/activate  # On Windows: myenv\Scripts\activate
pip install -r requirements.txt

Create a .env file in Backend with:

DB_PASSWORD= "your_db_password"
SERP_API_KEY = "your_serpapi_key"
DATABASE_URL = "postgresql://username:password@localhost/travel_ai"
GROQ_INTENT_KEY = "your_groq_key"
GROQ_API_KEY = "your_groq_key"
GEMINI_API_KEY = "your_gemini_key"
PINECONE_API_KEY = "your_pinecone_key"

Run the server:

cd itinerary_module
python main.py

2. Agent Backend

cd Backend/agent_module
python -m venv myenv
source myenv/bin/activate  # On Windows: myenv\Scripts\activate
pip install -r requirements.txt

Create a .env file in agent_module with:

DB_PASSWORD= "your_db_password"
DATABASE_URL= postgresql://username:password@localhost/travel_ai

SERP_API_KEY = "your_serpapi_key"

# API keys for Groq
GROQ_API_KEY= "your_groq_key"
GROQ_API_KEY1= ... # add more keys if needed

# twilio credentials
TWILIO_SID= your_twilio_sid
TWILIO_AUTH_TOKEN= your_twilio_auth_token
TWILIO_WHATSAPP_NUMBER=whatsapp: your_twilio_whatsapp_number
TWILIO_CALLER_ID= your_twilio_caller_id

# Cloudinary credentials
CLOUDINARY_CLOUD_NAME= your_cloudinary_cloud_name
CLOUDINARY_API_KEY= your_cloudinary_api_key
CLOUDINARY_API_SECRET=your_cloudinary_api_secret

# Eleven Labs API Key and Agent ID
ELEVEN_API_KEY= your_elevenlabs_api_key
ELEVEN_AGENT_ID= your_eleven_agent_id
ELEVEN_PHONE_NUMBER_ID= your_eleven_phone_number_id

# Bluesky credentials
BLUESKY_HANDLE= your_bluesky_handle
BLUESKY_APP_PASSWORD=your_bluesky_app_password

To get proper campaign insights we need to seed the database with some dummy data. Run the seeding script:

python seed_data.py

Run the server:

python main.py

3. Frontend

cd frontend
npm i
npm run dev

The product will now be accessible at:

http://localhost:3000

Startup Order

For smooth execution:

1. Start PostgreSQL
2. Run itinerary_module
3. Run agent_module
4. Start frontend

---

🧠 Itinerary Intelligence Module – Capabilities

• Natural language trip generation
• Live flight & hotel integration (SerpAPI)
• Combination engine (Flight × Hotel)
• Multi-objective scoring
• Risk & margin engines
• Multi-agent debate reasoning
• What-if simulation endpoint
• Day-wise structured experience planner
• Contextual conversational query bot (LangGraph-based)

📈 Campaign Intelligence Module – Capabilities

• Destination trend analysis
• Smart customer targeting (RFM-based)
• AI-generated campaign blueprints
• AI-powered creative generation
• Parallel WhatsApp automation (ThreadPoolExecutor)
• Agent autocalling for real-time campaign adjustments
• Bluesky integration for automated social media posting
• Communication logging & analytics

---

🛠 Tech Stack

Backend

• Python
• Flask
• SQLAlchemy ORM
• PostgreSQL

AI & Intelligence

• Google Gemini API
• Groq API
• LangGraph
• Pinecone (Vector DB)
• Multi-agent deterministic reasoning layer

External Intetgrations

• Twilio (WhatsApp)
• Cloudinary (Creative Hosting)
• ElevenLabs (Voice Synthesis)
• Bluesky (Social Media)

Scheduling

• APScheduler (Python)

Frontend

• Next.js

---

🚀 Why This Platform Is Different

Traditional travel systems:

• Static listings
• Manual marketing
• No explainability
• No profitability visibility

TBOAnalytica:

• AI-driven acquisition
• AI-driven decision optimization
• Margin intelligence
• Risk transparency
• Simulation modeling
• Structured campaign automation
• Conversational copilot
• Day-wise experiential planning

It merges Growth Intelligence + Decision Intelligence into one unified ecosystem.

Vision

This project demonstrates how TBO can evolve into:

An AI-Powered Travel Growth & Decision Operating System

Combining:

• Demand Intelligence
• Marketing Automation
• Booking Optimization
• Agent Profitability Intelligence
• Customer Experience Planning

All within one cohesive AI-driven platform.

Video Demo