EnergyWise - Electricity Bill Analyzer

EnergyWise is a web application that uses artificial intelligence to analyze electricity bills, helping customers better understand their consumption, identify possible errors, and receive personalized advice to save money.

Key Features

• Bill Analysis: Upload your electricity bill in PDF or image format (JPG, PNG) and get a detailed analysis.
• Bill Comparison: Compare the current bill with a previous one to identify variations in consumption and costs.
• Anomaly Detection: Identify possible errors or anomalies in bills.
• Personalized Advice: Receive suggestions on how to save energy and reduce costs.
• Virtual Assistant: Chat with an AI assistant to get answers to your questions about the bill.
• Responsive Interface: Works on desktop and mobile devices.

Technologies Used

• Backend: Flask (REST API), Python
• Frontend: HTML, CSS, JavaScript, Bootstrap 5
• AI/ML: AWS Strands Agents with Claude 3.5 Sonnet on Amazon Bedrock
• OCR: Pytesseract for text extraction from images and PDFs
• Alternative Interface: Streamlit for an alternative version of the application

System Requirements

• Python 3.8+
• Tesseract OCR
• Poppler (for PDF conversion)
• AWS CLI configured with access to Amazon Bedrock

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/bolletta-analyzer.git
   cd bolletta-analyzer

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Install Tesseract OCR:

   • macOS: brew install tesseract tesseract-lang
   • Ubuntu/Debian: sudo apt-get install tesseract-ocr tesseract-ocr-spa tesseract-ocr-eng
   • Windows: Download the installer from GitHub

5. Install Poppler:

   • macOS: brew install poppler
   • Ubuntu/Debian: sudo apt-get install poppler-utils
   • Windows: Download from poppler-windows

6. Configure environment variables:

   cp .env.example .env

   Edit the .env file with your AWS credentials and other configurations.

Usage

Flask Version

1. Start the Flask application:

   python app.py

2. Open your browser and go to http://localhost:8080

Streamlit Version

1. Start the Streamlit application:

   streamlit run streamlit_app.py

2. The browser will automatically open with the Streamlit application.

Project Structure

bolletta-analyzer/
├── app/                    # Flask application
│   ├── __init__.py         # Flask app initialization
│   ├── routes.py           # API routes
│   ├── services/           # Business logic services
│   │   ├── bill_analyzer.py    # Bill analysis with direct Bedrock (backup)
│   │   ├── strands_agent.py    # Strands Agent implementation
│   │   └── document_processor.py  # Document processing and OCR
│   ├── static/             # Static files (CSS, JS, images)
│   └── templates/          # HTML templates
├── venv/                   # Python virtual environment
├── .env                    # Environment variables
├── app.py                  # Flask application entry point
├── streamlit_app.py        # Alternative Streamlit application
└── README.md               # Documentation

AWS Configuration

The application uses AWS Strands Agents with Amazon Bedrock for bill analysis. Make sure you have:

1. An AWS account with access to Amazon Bedrock
2. Configured AWS credentials locally (~/.aws/credentials)
3. Set the correct region in the .env file (default: us-east-1)
4. Access to the Claude 3.5 Sonnet model on Amazon Bedrock

License

This project is released under the MIT license.

Contact

For questions or support, contact [anespo28@gmail.com].