ARCHITECTING INTELLIGENT AGENTS IN AZURE

Architecting Intelligent Agents in Azure takes you from your first interaction with an AI agent to deploying a fully orchestrated, production-ready system that can reason, remember, collaborate, and act. Using the Microsoft Agent Framework and Python, this book walks you through every engineering layer behind modern agentic systems, from grounding and memory to tools, semantic recall, and autonomous action. Across the chapters, you’ll build Thain, an Azure-native feedback agent that evolves into a multi-agent system with the ability to retrieve knowledge, coordinate tasks, and refine its performance. With a balanced mix of architecture, hands-on code, and cloud patterns, you’ll learn how Azure Cosmos DB, Azure AI Search, Azure Monitor, and serverless components come together to form intelligent, self-improving enterprise agents. Whether you’re an AI engineer, Azure developer, or solution architect, this book offers a practical, end-to-end guide to building agents that can reason, recall, collaborate, and grow over time. What You Will Learn: Build a fully functioning agent using Python and the Microsoft Agent Framework Implement reasoning loops, short-term memory, and tool-based actions Add persistent memory using Azure Cosmos DB and semantic recall using Azure AI Search Create safe and governed agents with telemetry, observability, and policy enforcement