aaronjjean / pani-genie

Backend of the pani-genie chatbot

About Pani-Genie

Pani-Genie is an innovative chatbot application based on the Ayushma framework designed by the designed to provide the users of Indiawaterportal with detailed information and assistance related to water management and related topics. Leveraging advanced AI technologies, Pani-Genie integrates OpenAI's GPT-3.5 turbo for natural language processing and Pinecone for vector database management to offer accurate and relevant information. The chatbot is built to support students, scholars, and experts in accessing a vast repository of articles and data, making it a valuable resource for educational and professional purposes.

1. Download VS Code

   • Download VS Code
   • Open the setup
   • Follow the installation instructions

   ()

2. Download Git

   • Download Git for Windows
   • Select "Clone a repository"
   • Enter the link to the backend: https://github.com/coronasafe/ayushma.git
   • OR download the backend_ay folder from the drive and continue.

3. Download a Remote Server (WSL - Ubuntu)

   • WSL Tutorial
   • Follow the instructions to install WSL
   • Open a remote server

4. Enable Windows Subsystem for Linux

   • Search for "Turn Windows features on or off"
   • Select "Windows Subsystem for Linux" and turn it on
   • Restart the computer when prompted
   • Add the distro (Ubuntu) and install it
   • Launch Ubuntu and set a username and password

   

5. Connect to the Distro

   • You should see WSL: Ubuntu
   • Go to the /mnt directory to access files on Windows

   

Setting Up the Backend

1. Open a New Folder in VS Code

   • Select the file location in the remote server
   • Use Ctrl + to open the terminal

   

2. Download Docker

   • Download Docker
   • Install using the installer
   • Continue without signing in and restart if the engine stopped
   • Make sure the Docker engine is running and restart if needed

3. Install the make Command

   • Install make using the appropriate command for your environment

   

4. Run the make up Command

   • Ensure Docker is running
   • The make command will initiate the required installations
   • Once all dependencies are installed and containers are started, run the make down command to stop all containers

   

5. Set Environment Variables

   • In the ayushma/.envs/.local/django file, append the following:

     AI_NAME="Your api"
     OPENAI_API_KEY=sk-3axxxxxxxxxlOG
     PINECONE_API_KEY=43xxxxxxx2
     PINECONE_ENVIRONMENT=gcp-starter
     PINECONE_INDEX=default
     CURRENT_DOMAIN=http://localhost:8000

   • Set the environment variables:
     • Find the OpenAI key:
       • Login to OpenAI API
       • Go to the API-Keys section and create a new key
       • Save it to a notepad
     • Login to Pinecone
       • Go to the API-keys section and copy the default key
       • Set the environment variable
     • Store these keys in the variable and set the Pinecone index and environment according to the index available in the Pinecone page
   • Run the make up command

Setting Up the Frontend

1. Install Node.js

   • Download Node.js

2. Install Yarn

   • Run the command:

     npm install --global yarn

   • Once Yarn is installed, run:

     yarn dev

   • If issues arise, try changing the execution policy from Restricted

   

Adding a Superuser

1. Access the Entrypoint File

   • Run the command:

     docker exec -it ayushma-django-1 /entrypoint bash

   • This accesses the entrypoint file where you can add the superuser

   

2. Create the Superuser

   • Run the command:

     python manage.py createsuperuser

   • Enter the credentials

3. Add a New Project in the Admin Panel

   • Go to localhost:8000/admin
   • Add a new project with a prompt
   • Fill in all required details

   

4. Set the Project as Default

   • Select the "is default" option

Login to the Application

1. Go to localhost:3000 for Login

   • Login with the credentials:
     • Email: [email protected]
     • Password: Test@2024
   • Login instructions:here

2. Setup a Local Server (if needed)

   • Change the URL from localhost:8000 to https://ipv4address:8000 in the backend and frontend