Azure AI Agents Orchestration Playbook

A comprehensive collection of tutorials for building production-ready Azure AI Agents, from basic conversations to complex multi-agent orchestration systems with voice capabilities and external API integrations.

Note: Check out our animated PowerPoint presentation for an overview: Azure AI Agents Orchestration Playbook Presentation

Tutorial Structure

Foundational Concepts

• 01-agent-basics - Core Azure AI Agents fundamentals using Foundry SDK and Semantic Kernel
• 02-agent-custom-functions - Extend agent capabilities with custom functions and plugins

Multi-Agent Systems

• 03-orchestrated-agents - Coordinate multiple agents for complex workflows
• 04-orchestrated-agents-with-tools - Integrate external APIs and Azure Logic Apps
• 05-orchestrated-agents-with-custom-openapi-tools - Connect agents to custom FastAPI services
• 06-magentic-one-orchestration - Advanced orchestration with Magentic-One framework

Advanced Capabilities

• 07-voice-orchestration - Voice-enabled agents with real-time audio processing
• 08-agent-routing - Hierarchical routing design pattern for multi-agent systems

Getting Started

Prerequisites

• Azure subscription with AI services
• Azure AI Foundry project with deployed models
• Python 3.8+ with Jupyter support

Quick Setup

1. Clone the repository

   git clone https://github.com/Azure-Samples/azure-ai-agents-playbook.git
   cd azure-ai-agents-playbook

2. Configure environment

   • Update .env file with your Azure AI project details
   • Install dependencies: pip install -r requirements.txt

3. Start learning

   • Begin with 01-agent-basics for fundamentals
   • Progress through tutorials in numerical order

For detailed instructions, please follow the step by step guide here .

Learning Path

Recommended progression:

1. Basics → Learn core Azure AI Agents concepts and development patterns
2. Functions → Add custom capabilities to extend agent functionality
3. Orchestration → Coordinate multiple agents for complex workflows
4. Tools → Integrate external services and automation workflows
5. Custom APIs → Connect agents to your own business services
6. Advanced → Master sophisticated orchestration frameworks
7. Voice → Add voice interaction capabilities

What You'll Build

By completing this playbook, you'll master:

• Conversational AI Agents with persistent context and memory
• Multi-Agent Systems that collaborate intelligently
• Tool-Enhanced Agents connected to external APIs and services
• Voice-Enabled Agents with real-time audio processing
• Enterprise Workflows combining AI with business automation
• Production Patterns for scalable and reliable agent systems

🔧 Key Technologies

• Azure AI Agents - Core agent development platform
• Azure AI Foundry SDK - Official Microsoft agent framework
• Semantic Kernel - Advanced orchestration and plugin system
• Azure Speech Services - Voice recognition and synthesis
• Azure Logic Apps - Workflow automation and integration
• FastAPI - Custom API development and OpenAPI integration

📋 Environment Configuration

Each tutorial folder contains specific setup instructions. The main environment variables needed:

• PROJECT_ENDPOINT - Your Azure AI Foundry project endpoint
• MODEL_DEPLOYMENT_NAME - Your deployed AI model name
• AZURE_OPENAI_* - Azure OpenAI configuration for Semantic Kernel
• AZURE_VOICE_* - Speech services for voice tutorials

🎪 Tutorial Highlights

Real-world scenarios you'll implement:

• Financial assistants analyzing bank transactions
• Research agents conducting web searches and fact-checking
• Creative content pipelines with multi-agent collaboration
• Email automation through voice commands
• Currency exchange agents with live API data
• Complex business workflows with AI decision-making

🔗 Additional Resources

• Azure AI Agents Documentation
• Semantic Kernel Documentation
• Azure AI Foundry

---

Ready to build the future of AI agents? 🚀🤖