This project demonstrates how to use the Microsoft Agent Framework to build AI-powered workflows for data analysis and reporting.
- Python 3.12 or higher
- Windows OS (if you are using the provided batch file)
Run the setup batch file to create and configure the virtual environment:
setup_venv.batThis will:
- Create a virtual environment named
agentframework - Activate the environment
- Install all required dependencies from
requirements.txt
Create a .env file in the project root with your Azure OpenAI credentials:
API_KEY=your_azure_openai_api_key
AZURE_AI_MODEL_DEPLOYMENT_NAME=your_deployment_name
AZURE_AI_PROJECT_ENDPOINT=your_endpoint_url
AZURE_AI_API_VERSION=your_api_versioncall agentframework\Scripts\activatepython main.pyThis will:
- Analyze the sample sales data from
sample_sales_data.csv - Generate a comprehensive sales report using a multi-agent workflow
- Display the conversation between the data analyzer and report writer agents
The main.py script demonstrates a sequential agent workflow:
- Data Analyzer Agent: Reads CSV data and calculates key statistics (revenue, top performers, etc.)
- Report Writer Agent: Transforms the data into a stakeholder-friendly report
This is useful for automating daily sales summaries for managers, executives, and team leads in various contexts (restaurants, retail, e-commerce, etc.).
.
├── agentframework/ # Virtual environment (created by setup_venv.bat)
├── main.py # Main sales report workflow demo
├── sample_sales_data.csv # Sample data for testing
├── requirements.txt # Python dependencies
├── setup_venv.bat # Environment setup script
└── README.md # This file
When you're done working, deactivate the virtual environment:
deactivate