Airline Customer Support Bot

Overview

This repository contains the code for an Airline Customer Support Bot designed to assist with customer queries efficiently and effectively.

Tools Used

• Langchain
• ChromaDB (vector store)
• Ollama (Llama 3.1 AI model)
• MySQL (chat history)
• Flask

Setup Instructions

Ensure you have Ollama running:

ollama serve

Set up a virtual environment:

python -m venv venv

Install the necessary libraries:

pip install -r requirements.txt

Modify the `.env` file with your specific information.

Create the `messages` table using create_messages_table.sql:

Make sure you have db and pdf directories in the same directory.

Execute the `app.py` file to start the application:

python app.py

API Endpoints

/ai

• Description: Ask LLM
• Chat History: None
• Context from PDFs: None

/airbot

• Description: Ask LLM
• Chat History: Includes a maximum of 6 past messages
• Context from PDFs: Yes

/pdf

• Description: Upload PDFs to the vector store

Contribution

Feel free to open issues or submit pull requests if you have suggestions or improvements.