Unimelb Handbook Chatbot

This is a chatbot integrated with OpenAI and Vanna AI, connected through a Chrome extension that provides users with real-time assistance based on the UoM Student Handbook.

Try the Chatbot

Chrome Extension Version
1. Download the package from Latest Realease page and unzip it.
2. Load this directory in Chrome as an unpacked extension.
3. Open the extension's side panel by clicking the hoverball on the bottom right of your screen to open the side panel.
Website Version (Not maintaining)

Deployment guideline

Prerequisites

OpenAI: Obtain an API key from the OpenAI website. This key is required for integrating OpenAI services.
Vanna AI Keys: Obtain an API key from Vanna AI. This key is required for interacting with Vanna AI services.
PostgreSQL Database: Set up your own PostgreSQL database to store the necessary data. You can use cloud providers like AWS RDS, Google Cloud SQL, or any other preferred method to host your PostgreSQL database.

Step 1: Clone the repository

git clone https://github.com/may-bee/it-project2024-sm-2.git

Step 2: Environment Setup

Install Python
Ensure Python 3.x is installed on your system. You can verify by running:
```
python3 --version
```
Create a Virtual Environment
Set up a virtual environment to isolate the dependencies:
```
python3 -m venv venv
```
Activate the Virtual Environment
For Windows:
```
venv\Scripts\activate
```
For Linux or macOS:
```
source venv/bin/activate
```
Install Dependencies
Install the necessary dependencies from requirements.txt:
```
pip install -r requirements.txt
```

Step 3: Vercel CLI Installation

Install Node.js and NPM
Ensure that Node.js and npm are installed on your machine, as Vercel requires npm. You can check the versions with:
```
node -v
npm -v
```
Install Vercel CLI
To install the Vercel command-line tool globally, run:
```
npm install -g vercel
```

Step 4: Build Procedures

Vercel automatically detects and builds your project when deploying. However, to ensure everything goes smoothly, set up the necessary build command.

Create a Start Script
Add a start command to app.py if not already present:

if __name__ == '__main__':
   from os import environ
   app.run(debug=False, port=int(environ.get('PORT', 5000)), host='0.0.0.0')

Ensure Dependencies
Make sure all necessary Python dependencies are included in requirements.txt.

Step 5: Deploying to Vercel

Log in to Vercel
Use the Vercel CLI to log in:
```
vercel login
```
Initialize Deployment
Navigate to your project directory and initialize Vercel deployment:
```
vercel
```

Step 6: Environment Variables

Add Environment Variables
Use the .env.sample as a reference for necessary environment variables. You can add these to Vercel's configuration:
```
vercel env add VARIABLE_NAME VARIABLE_VALUE
```
You can also use the Vercel Dashboard to manage environment variables.

Step 7: Push to Production

Test the Deployment
After the initial deployment, Vercel provides a unique staging URL where you can test your API to ensure it works as expected.
Promote to Production
If everything looks good on the staging URL, you can promote the app to production using:
```
vercel --prod
```

Vercel will handle the build and deployment processes and provide you with an updated production link.

System Architecture Overview

User Interaction: Users interact through a Chrome extension that pops up when they access the UoM Student Handbook in their browser.
Chrome Extension: This extension, comprising HTML, JavaScript, and manifest files, triggers the UoM Handbook Chatbot popup.
Chatbot API: The Chrome extension sends user questions to the Flask Python-based chatbot API.
Vanna AI & OpenAI Integration: The chatbot API sends user questions to Vanna AI for query processing. If natural language understanding or additional processing is needed, OpenAI is used to generate responses.
Supabase: The data used by Vanna AI is stored in Supabase, which contains information scraped from the University of Melbourne Student Handbook.

IyXuan23 / ITProject2024SM2

readme