Pharma Knowledge Assistant💊

Project Overview

The Pharma Knowledge Assistant is a comprehensive tool designed to assist users with pharmaceutical-related queries. It leverages a combination of Agentic-Retrieval Augmented Generation, a pharmaceutical knowledge base, and a user-friendly interface to provide detailed, accurate, and helpful information about medications, their alternatives, dosages, side effects, and more.

Project Structure

The project is organized into several key components:

• src/: Contains the main source code for the project.
  • main.py: The main entry point for the Streamlit application.
  • data-processing.py: Script for processing and embedding pharmaceutical data into ChromaDB.
  • supervisor_arch.ipynb: Jupyter notebook for an step-by-step workflow the supervisor architecture and agents with the agentic workflow.
• datasets/: Contains JSON files with pharmaceutical data.
• chroma_db/: Directory for ChromaDB storage.
• requirements.txt: Lists all the dependencies required for the project.
• web_scrapper.py: Script for scraping pharmaceutical data from specified URLs.
• actual_dataset.zip: Compressed file containing the raw dataset.
• pes_esa_hackathon.pdf: PDF document that contains the problem statement.

Setup Instructions

Prerequisites

Ensure you have the following installed on your system:

• Python 3.12.7 or later
• pip (Python package installer)

Installation

1. Clone the repository:

   git clone 
   cd pharma-knowledge-assistant

2. **Create a virtual environment:

   python -m venv venv
   source venv/bin/activate 

3. Install the required packages:

   pip install -r requirements.txt
   

Configuration

1. Process the document folder and store populate the database, with the vectorised chunks of the dataset.

   python src/data-processing.py
   

2. Configure API keys: Update the api_key in main.py and supervisor_arch.ipynb with your OpenAI API key.

   llm = ChatOpenAI(
   temperature=0,
   api_key="your_openai_api_key",
   model="gpt-4o-2024-11-20"
   )
   

Running the Application

1. Start the Stremlit applicaton

   streamlit run src/main.py
   

2. Access the application:

Open your web browser and navigate to http://localhost:8501 to interact with the Pharma Knowledge Assistant.

Usage

• Ask Questions: Use the chat input to ask any pharmaceutical-related questions.
• Get Recommendations: Receive medication recommendations based on symptoms or conditions.
• Find Alternatives: Get safe alternatives for medications.
• Summarize Products: Obtain concise summaries of pharmaceutical products.

Supervisor Agentic Architecture

This diagram demonstrates how agents interact with the pharmaceutical knowledge base and respond to user queries.

Screenshots

Data-Processing

UI

Simple RAG

Summarizer

Integration and Complex Scenario Analysis