Building AI Agent with Semantic Kernel

In this comprehensive session, Johnny Hooyberghs from the Microsoft Zero to Hero Community demonstrates how to build your own AI agent using Semantic Kernel. The session starts with an introduction by the host, emphasizing the importance of the Microsoft Learn Community and the MS Build event. Johnny, a seasoned software developer and Microsoft MVP, dives into the core topic, explaining the different levels of AI interaction from simple chatbots to fully autonomous agents. He highlights the capabilities of Semantic Kernel, a framework that integrates AI functionalities into .NET applications using plugins, personas, and planners. Johnny provides several live demos, showcasing how to configure Semantic Kernel to interact with Azure OpenAI and create intelligent AI agents. He demonstrates the use of native functions to enrich AI responses with real-time data and how to maintain chat history for contextual conversations. Johnny also presents a real-world application, a home automation dashboard, where he integrates a co-pilot to answer questions and perform tasks based on real-time data. Additionally, he showcases a game called C Wars, where students can write C code to control robots in an arena, enhanced with AI-generated scripts using Semantic Kernel. The session concludes with a Q&A, addressing various questions about Semantic Kernel’s capabilities and its comparison with other AI frameworks. Johnny’s detailed explanations and practical demos provide a thorough understanding of how to leverage Semantic Kernel for building advanced AI agents.