📋 Workshop Prerequisites

**You will need:** - **Azure tenant with active subscription** - Use a demo/development tenant (NOT your production tenant) - You likely already have access to an Azure tenant - just ensure your subscription is active **Important:** All resources will be deployed to a **new resource group** that you'll create during the workshop. This resource group can be easily deleted and recreated, allowing you to repeat the training as many times as you'd like without affecting existing resources.

**Good News!** 🎉 Our deployment process handles most permission assignments automatically. Here's what the solution uses: | Role | Used For | Assigned To | |------|----------|-------------| | **Contributor** | Creating and managing Azure resources | Your user account (required) | | **User Access Administrator** | Assigning managed identities | Your user account (required) | | **Cognitive Services Contributor** | Azure OpenAI setup | Your user account (required) | **Important:** You need **Contributor** and **User Access Administrator** roles on the subscription or resource group before starting. The deployment will automatically configure other permissions like: - ✅ Managed identity for Container App - ✅ Storage account role assignments - ✅ Azure OpenAI access for the app **To check your current permissions:** ```bash az role assignment list --assignee $(az account show --query user.name -o tsv) \ --query "[].roleDefinitionName" -o table ``` **Don't have the right permissions?** Ask your Azure administrator to grant you **Contributor** and **User Access Administrator** roles on the resource group you'll use for this workshop.

The workshop uses these Azure services: **Check Azure OpenAI access:** ```bash az cognitiveservices account list --query "[?kind=='OpenAI']" -o table ```

**NOT REQUIRED for this workshop!** The workshop uses **Azure Developer CLI (azd)** for deployment, which performs cloud-based container builds. You do **NOT** need Docker Desktop installed. **Advanced Option:** If you want to test containers locally or build images manually, you can optionally use: - **Docker CLI** (command-line only): Included with Docker Desktop or standalone - **Docker Desktop**: [Download here](https://www.docker.com/products/docker-desktop/) **Command-line build example (optional):** ```bash # Only needed if you want to build/test locally docker build -t myapp . docker run -p 8080:8080 myapp ```

**Important Note on Speech Features:** This workshop solution uses **browser-based Web Speech API** for voice recognition, NOT Azure Cognitive Services. This means: - ✅ Works great on desktop browsers (Chrome, Edge, Safari) - ⚠️ Mobile browsers support text chat only (no voice) - 💰 No additional Azure speech costs - 🎓 Perfect for learning and prototyping In production, you could upgrade to Azure Cognitive Services Speech SDK for mobile support and enhanced features.

### Included Dependencies The workshop uses these Python packages (automatically installed during deployment): **Web Framework:** - Flask 3.1.0 - Web application framework - Flask-SocketIO 5.4.1 - Real-time communication - Flask-Login 0.6.3 - User authentication - msal 1.31.1 - Microsoft authentication **Azure SDKs:** - azure-identity - Azure authentication - azure-ai-inference - AI model integration - azure-data-tables - Table Storage for analytics - agent-framework-azure-ai - Microsoft Agent Framework **Additional Tools:** - python-dotenv - Environment configuration - requests - HTTP library **You don't need to install these manually!** The `requirements.txt` file is included and azd will handle installation during deployment.