This repository contains an AI-powered assistant for analyzing and answering questions about personal data files, implemented in two different ways.

• Implementations
  • Jupyter Notebook
  • Python Scripts
• Getting Started
• Usage
• Contributing

This implementation is contained in Chatbot_personal_data.ipynb and uses:

• Hugging Face models for natural language processing
• LangChain for document processing and question answering
• Gradio for the user interface

• Uses the facebook/bart-large-cnn model for text generation
• Implements document chunking and embedding using sentence-transformers/all-MiniLM-L6-v2
• Creates a simple UI with Gradio for file upload and question answering

This implementation is split across multiple Python files and uses:

• IBM Watson Machine Learning and WatsonxLLM for language modeling
• LangChain for document processing and question answering
• Flask for the backend server
• HTML, CSS, and JavaScript for the frontend

• worker.py: Core logic for document processing and question answering
• server.py: Flask server to handle API requests
• Frontend files (HTML, CSS, JS) for the user interface

• Uses the meta-llama/llama-2-70b-chat model through IBM Watson Machine Learning
• Implements advanced document chunking and retrieval methods
• Provides a web-based interface for file upload and chatting

1. Clone this repository: git clone https://github.com/yourusername/ai-assistant-personal-data.git cd ai-assistant-personal-data

2. Install the required dependencies: pip install -r requirements.txt

3. For the Jupyter notebook implementation:

• Open and run Chatbot_personal_data.ipynb in a Jupyter environment

4. For the Python script implementation:

• Set up your IBM Watson Machine Learning credentials
• Run the Flask server:

  python server.py
  

• Open the provided HTML file in a web browser to access the interface

1. Upload a PDF document containing personal data
2. Ask questions about the content of the document
3. Receive AI-generated answers based on the document's content

In this Demo, I uploaded my resume to the app. Watch the Demo Video

Contributions to improve either implementation are welcome. Please feel free to submit issues or pull requests.

This project demonstrates two different approaches to building an AI assistant for personal data analysis, showcasing various technologies and implementation methods.