Technical Challenge: LLM Airline Policy App

Overview

Welcome the Flight Center AI Center Of Excellence technical challenge. We're excited about your interest in joining us using AI to revolutionize the travel industry.

The purpose of this challenge is to assess your skills in using Python and Large Language Models (LLMs).

Challenge Description

Your task is to create a small application where a user can ask questions about the policy of an airline. The application should query the airline documents to retrieve relevant information based on the user questions. The application should leverage an LLM for understanding and answering questions and a vector database to store and retrieve document embeddings efficiently.

We've included a collection of airline policies. Your task is to create an application where users can ask questions and get answers for queries like the following:

1. Can my pet travel with me on the plane on Delta?

2. I have three kids 2, 8 and 10 years old and I am traveling with them on a United flight, what are the rules for children traveling?

3. What is the baggage policy for American Airlines?

4. My wife is 8 months pregnant, can she travel on a Delta flight?

Objectives

Build a chatbot application that can answer questions about airline policies. The application should have the following features:

1. User Interface: A simple web interface where users can input questions and get answers.
2. LLM Integration: Use a pre-trained LLM model to understand and answer questions.
3. Document Processing: Extract text from the airline policy documents.
4. Vector Database: Store document embeddings in a vector database for efficient similarity search.

Provided Policies

The airline policy documents are provided in PDF and markdown format. You will need to process these documents to extract the text before processing.

Deliverables

• Code hosted on a private GitHub repository.
  • Fork this repository to your GitHub account.
  • Push your developed code to your fork.
  • Email us the link to your private repository when you are ready for review.
• A README file documenting:
  • How to set up and run your application.
  • A brief explanation of your design choices and technologies used.
  • Any challenges you encountered and how you resolved them.

LLM Model

You can use any pre-trained LLM model of your choice. We recommend using OpenAI's GPT models, an API Key will be provided to you.