Project Documentation

Table of Contents

1. Introduction
2. Backend
3. Frontend
4. Setup
5. License

Introduction

Simple QA Chatbot interface powered by deepset/roberta-base-squad2 and haystack pipeline

Demo 🚀

Demo Video ⭐

Demo images ⭐

TODO List

• Upload document and Process
• Play Audio button beside respective answer
• Make the store scalable (Elastic Search Document Store)
• Fine Tune the Reader
• Implement user authentication
• Add a Databse to store chats
• Enhance UI/UX design

Backend

Technologies Used

• Python
• FastAPI
• Haystack

Frontend

Technologies Used

• React.js, Vite
• Axios
• TailwindCSS

Setup

To set up the project locally, follow these steps:

Backend

1. Navigate to the backend folder:

cd backend

2. Create a virtual environment:

python -m venv venv

3. Activate the virtual environment:

• On Windows:

  venv\Scripts\activate

• On macOS/Linux:

  source venv/bin/activate

4. Install dependencies:

pip install -r requirements.txt

5. Start the server:

uvicorn main:app --reload

Frontend

1. Navigate to the frontend folder:

cd frontend

2. Install dependencies:

npm install

3. Start the development server:

npm run dev

License

This project is licensed under the MIT License. See the LICENSE file for details.