NeuroPharmaQuantix Omniscient Architecture with Recursive Knowledge Synthesis v2.7.3-Helios

🧬 Advanced Pharmaceutical Knowledge Assistant powered by Neural Embeddings and GroqAI 🤖

🌟 Overview

NPQ-OARKS is a sophisticated pharmaceutical knowledge assistant that leverages neural embeddings and large language models to provide accurate, context-aware information about pharmaceutical products. Built using state-of-the-art technologies, it offers an intuitive chat interface for accessing detailed pharmaceutical knowledge.

🏗️ Architecture

flowchart TB
    subgraph Data Pipeline
        A[Web Scraper] -->|JSON Files| B[Data Processing]
        B -->|Text Chunks| C[Neural Embeddings]
        C -->|Vector Store| D[ChromaDB]
    end
    
    subgraph Core System
        E[GroqAI LLM] -->|Query Processing| F[Knowledge Synthesis]
        D -->|Context Retrieval| F
        F -->|Response Generation| G[Agent System]
    end
    
    subgraph User Interface
        H[Streamlit Frontend] -->|User Query| G
        G -->|Enhanced Response| H
    end
    
    style Data Pipeline fill:#1e2a4a,color:#fff
    style Core System fill:#2a1e4a,color:#fff
    style User Interface fill:#4a1e2a,color:#fff

Loading

✨ Features

• Advanced Neural Processing: Utilizes HuggingFace's all-MiniLM-L6-v2 for state-of-the-art text embeddings
• Intelligent Context Synthesis: Implements recursive character splitting for optimal text processing
• GPU-Accelerated Performance: Leverages CUDA acceleration for enhanced processing speed
• Persistent Knowledge Store: ChromaDB-based vector storage for efficient information retrieval
• Interactive UI: Sleek, animated interface with realtime response visualization
• Multi-Tool Integration:
  • Product Knowledge Base
  • Smart Summarizer
  • Intelligent Recommender

🚀 Quick Start

1. Clone the Repository

git clone https://github.com/yourusername/NeuroPharmaQuantix-OARKS.git
cd NeuroPharmaQuantix-OARKS

2. Environment Setup

python -m venv env
source env/bin/activate  # On Windows: env\Scripts\activate
pip install -r requirements.txt

3. Configure Environment Variables

# Create .env file
GROQ_API_KEY=your_api_key_here

4. Run the Application

streamlit run try1.py

💻 Technical Implementation

Data Pipeline

• Web Scraping: Custom scraper for pharmaceutical data acquisition
• Text Processing: RecursiveCharacterTextSplitter with 500-token chunks
• Embedding Generation: GPU-optimized batch processing
• Vector Storage: Persistent ChromaDB implementation

Core Components

# Key System Components
├── ChatGroqManager        # LLM Interface Management
├── PharmaKnowledgeAssistant
│   ├── Embeddings Engine
│   ├── Vector Store
│   ├── Tools System
│   └── Agent Framework
└── UI System

🛠️ Development Stack

• Frontend: Streamlit with custom CSS animations
• Backend: Python with LangChain framework
• Database: ChromaDB for vector storage
• AI Models:
  • GroqAI llama-3.2-90b-vision-preview
  • HuggingFace all-MiniLM-L6-v2

🎯 Performance Optimization

• Batch processing for embedding generation
• CUDA acceleration for GPU systems
• Efficient memory management with periodic cache clearing
• Optimized chunk size for context retrieval

🎨 UI/UX Features

• Dark mode interface with radial gradient
• Animated message transitions
• Interactive chat bubbles
• Real-time response indicators
• Custom scrollbar implementation
• Responsive design elements

📊 System Architecture Components

graph TB
    A[Web Interface] -->|User Input| B{Router}
    B -->|Queries| C[LLM Processor]
    B -->|Searches| D[Vector Store]
    C -->|Context Request| D
    D -->|Embeddings| E[Neural Engine]
    C -->|Final Response| A
    
    style A fill:#2a3f5f,color:#fff
    style B fill:#5f2a3f,color:#fff
    style C fill:#3f5f2a,color:#fff
    style D fill:#2a5f3f,color:#fff
    style E fill:#5f3f2a,color:#fff

Loading

🔄 Setup Process Flow

sequenceDiagram
    participant U as User
    participant S as System
    participant DB as Database
    participant LLM as GroqAI
    
    U->>S: Initialize Application
    S->>DB: Check Existing Embeddings
    alt Embeddings Exist
        DB-->>S: Load Existing Data
    else No Embeddings
        S->>S: Generate New Embeddings
        S->>DB: Store Embeddings
    end
    S->>LLM: Test Connection
    LLM-->>S: Confirm Ready
    S->>U: Display Interface
    
    Note over U,S: System Ready for Queries

Loading

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

🙏 Acknowledgments

• MicroLabs USA for pharmaceutical data
• Groq team for AI infrastructure
• LangChain community for framework support

📸 Demo Screenshot

🔑 Important Notes for Users Environment Setup Before running the application, make sure to:

Create a .env file in the root directory with your API keys:

bashCopyGROQ_API_KEY=your_groq_api_key_here Dataset Options Using Medical Dataset (Default)

Download the pharmaceutical dataset:

Create a datasets folder in the root directory Inside it, create a microlabs_usa folder Run web_scrapper.py to populate the data Let the embeddings generate on first run

Using Custom Dataset You can adapt this architecture for any domain! Simply:

Prepare your data in JSON format with a similar structure Update the data loading path in try1.py Adjust the tool descriptions and prompts for your domain Let the system generate new embeddings for your data

The architecture is domain-agnostic and can be used for:

Legal document analysis Educational content delivery Technical documentation Customer support And much more!

License

This project is licensed under the MIT License.
© 2025 Siddharth

Made with 💊 by Siddharth Prakash