Speakers

This chapter explores techniques for monitoring and debugging multi-agent AI systems using Microsoft Agent Framework in .NET. It covers essential observability patterns, including configurable filters and comprehensive logging, to trace agent interactions and plugin calls.  Centralized auditing mechanisms provide transparency over execution flows and decisions made by each agent.The chapter also introduces Microsoft.Extensions.AI.Evaluation, a framework that empowers developers to systematically evaluate agent behaviors against expected outcomes to enhance reliability and robustness. Practical debugging strategies and best practices are discussed to identify bottlenecks, resolve errors, and optimize multi-agent workflows in production environments. With concrete C# examples and integration guidance, this chapter equips developers with the tools to gain operational insights and maintain trust in complex AI orchestration scenarios. Read More...