Practical AI Agents in C# - by Saravanan Ganesan (Paperback)

Book Synopsis

This book is a practical, hands-on guide to building intelligent AI agents on the modern Microsoft stack using Agent Framework 1.0. As Microsoft unifies Semantic Kernel and AutoGen into a single production-ready SDK, this book helps .NET developers move beyond scattered documentation and quickly start building real-world agentic applications in C#.

From spinning up a first agent against Microsoft Foundry to wiring tools through MCP and orchestrating multi-agent workflows, the book delivers a focused, implementation-first approach designed for developers who want to build--not just experiment. Through concise chapters packed with production-ready .NET 10 code, you will learn how to create capable agents with tools, plugins, middleware pipelines, memory systems, vector search, and graph-based orchestration patterns.

The book explores advanced enterprise scenarios, including human-in-the-loop approvals, AI workflow checkpoints, Azure AI Search integration, Teams and Microsoft 365 connectivity, observability with OpenTelemetry, semantic caching, model selection strategies, and secure deployment to Azure. Every chapter concludes with practical mini-projects that can be adapted immediately for real business applications.

Whether you are migrating from Semantic Kernel, experimenting with AI orchestration, or building scalable enterprise agents from scratch, this book provides a complete roadmap from prototype to production.

What You Will Learn:

• Create and configure AI agents using the Agent Framework 1.0 core abstractions and the Microsoft.Agents.AI NuGet package in C#
• Wire tools and external services into agents using MCP and the built-in .NET plugin model
• Compose multi-agent workflows using graph-based orchestration patterns -- sequential, concurrent, handoff, group chat, and Magentic-One
• Manage agent memory, context, and state for long-running and human-in-the-loop scenarios with Azure AI Search
• Deploy and observe production agents on Azure with OpenTelemetry, Application Insights, and Azure Functions hosting.

Who This Book Is For:

C# and .NET developers building AI-powered applications within the Microsoft and Azure ecosystem. Cloud engineers and solution architects on Azure adding agentic capabilities to existing .NET workloads

From the Back Cover

This book is a practical, hands-on guide to building intelligent AI agents on the modern Microsoft stack using Agent Framework 1.0. As Microsoft unifies Semantic Kernel and AutoGen into a single production-ready SDK, this book helps .NET developers move beyond scattered documentation and quickly start building real-world agentic applications in C#.

From spinning up a first agent against Microsoft Foundry to wiring tools through MCP and orchestrating multi-agent workflows, the book delivers a focused, implementation-first approach designed for developers who want to build--not just experiment. Through concise chapters packed with production-ready .NET 10 code, you will learn how to create capable agents with tools, plugins, middleware pipelines, memory systems, vector search, and graph-based orchestration patterns.

The book explores advanced enterprise scenarios, including human-in-the-loop approvals, AI workflow checkpoints, Azure AI Search integration, Teams and Microsoft 365 connectivity, observability with OpenTelemetry, semantic caching, model selection strategies, and secure deployment to Azure. Every chapter concludes with practical mini-projects that can be adapted immediately for real business applications.

Whether you are migrating from Semantic Kernel, experimenting with AI orchestration, or building scalable enterprise agents from scratch, this book provides a complete roadmap from prototype to production.

What You Will Learn:

• Create and configure AI agents using the Agent Framework 1.0 core abstractions and the Microsoft.Agents.AI NuGet package in C#
• Wire tools and external services into agents using MCP and the built-in .NET plugin model
• Compose multi-agent workflows using graph-based orchestration patterns -- sequential, concurrent, handoff, group chat, and Magentic-One
• Manage agent memory, context, and state for long-running and human-in-the-loop scenarios with Azure AI Search
• Deploy and observe production agents on Azure with OpenTelemetry, Application Insights, and Azure Functions hosting.

About the Author

Saravanan Ganesan is a Microsoft MVP in AI & IoT and Assistant Consultant at Tata Consultancy Services (TCS), based in Chennai, India. With over 12 years of hands-on experience in designing and architecting AI and IoT applications on the Microsoft Azure platform, he brings deep practical expertise to every solution he builds.

Saravanan specializes in multi-agent systems using Semantic Kernel and Microsoft Agent Framework, and has extensive experience with Azure AI services including Azure OpenAI, Document Intelligence, Content Safety, and Custom Vision. On the IoT side, he architects solutions across Azure IoT Hub, Event Hub, Digital Twins, and Device Provisioning Services -- bridging the worlds of intelligent agents and connected edge devices.

A sought-after international speaker, Saravanan has presented at conferences including .NET Conf Manila, AI Dev Tools Today, and Azure User Group Sweden, delivering sessions on topics ranging from Azure OpenAI chatbot development in .NET 8 and multi-agent systems with Semantic Kernel, to Industrial IoT 4.0 and Azure IoT Operations.

Beyond his technical work, Saravanan is a committed community builder and leader across the Microsoft developer ecosystem in India. He serves as the Organising Lead at Global AI Chennai, Chapter Lead for C# Corner Chennai, and Organiser of the Azure Developer Community Chennai -- three active communities where he regularly drives meetups, workshops, hackathons, and knowledge-sharing sessions that bring together hundreds of developers, architects, and AI enthusiasts across the region. He is also a mentor in the broader C# Corner global community, where he actively guides developers on their Microsoft AI and Azure journeys.

Outside of work, Saravanan genuinely enjoys the community side of things -- the meetups, the mentoring, the late-night GitHub threads. That same hands-on, no-shortcuts approach is what he's brought to this book.