Query-Me

The repository containing the code for an LLM powered chatbot for querying about your browser history and efficiently search through the sites you may have visited. Developed using Langchain and Streamlit. Uses PostgreSQL database to store the Chrome browser history.

Prerequisites:

• Python 3.x
• PostgreSQL (installed and running on localhost:5432 by default)
• A .env file with required environment variables. You can generate one using create_env_file.py.

Setting Up PostgreSQL Database:

Before you proceed with the backend and frontend setup, make sure you have a PostgreSQL database ready to use. If not, here's a brief guide to creating a new database and table.

1. Open a PostgreSQL terminal or use a GUI like pgAdmin.
2. Run the following SQL command to create a new database:

   CREATE DATABASE YOUR_DB_NAME;

Setting Up Backend:

Step 1: Clone the Repository

git clone https://github.com/SagarDollin/chatbot-personalData.git
cd chatbot-personalData

Step 2: Create a Virtual Environment (Optional but Recommended)

python -m venv venv
source venv/bin/activate   # On Windows, use 'venv\Scripts\activate'

Step 3: Install the Required Dependencies

pip install -r requirements.txt

Step 4: Create the .env File

Use the provided script to create a .env file with necessary configurations:

python create_env_file.py --api_key YOUR_OPENAI_API_KEY --db_pass YOUR_DB_PASSWORD --database YOUR_DB_NAME

Note: Replace placeholders like YOUR_OPENAI_API_KEY with actual values.

Step 5: Populate the Database

Download your chrome browserhistory.json file and save it in the Chrome folder at the root level.

Run db.py to populate the PostgreSQL database with a table called browserhistory.

python db.py

Step 6: Start the Backend Server

python server.py

You should see a message: OpenAPI key has been set, the server has been started!

Your backend server is now running on http://127.0.0.1:8001/.

You can view and send requests to the all the microservices on: http://127.0.0.1:8001/docs#

Using Postman with Backend:

1. Get Root:

Send a GET request to http://127.0.0.1:8001/ to get the server status.

2. Chat Query:

To chat with the bot, send a POST request to http://127.0.0.1:8001/chatDB/query_database with the following body:

{
  "prompt": "YOUR_PROMPT_HERE",
  "chat_history": "YOUR_CHAT_HISTORY_HERE"
}

Replace YOUR_PROMPT_HERE and YOUR_CHAT_HISTORY_HERE with your desired values.

3. Get Conversation Name:

To get a name for a conversation based on a user query, send a POST request to http://127.0.0.1:8001/conversation/get_name with the body containing your query as a string.

Running Frontend Application:

After setting up and running the backend server, you can run the Streamlit frontend as an example UI.

Step 1: Start the Frontend Application:

cd frontend
streamlit run frontendApp.py

You'll see a new browser window/tab opening with the ChatBot UI. Here, you can input messages and interact with the ChatBot, view previous chats, and more.

Conclusion:

This README provides steps to set up both the backend server and frontend application. Make sure the backend server is running when you're interacting with the frontend application. The server handles chat queries and also suggests conversation names based on user queries. The frontend serves as an example of how you can integrate and interact with the backend.