In the hands-on intensive course “AI Agents with .NET & Semantic Kernel,” your team will be guided through the creation of intelligent, enterprise-specific AI agents from scratch. Over three days, you will work through all relevant concepts and technologies to realize robust, customizable agents using Microsoft Semantic Kernel, .NET, and Retrieval-Augmented Generation (RAG) that store, link, and apply knowledge in production processes.
Request CourseOverview of agentic AI, LLM concepts, and RAG principles. Introduction to Microsoft Semantic Kernel: architecture, KernelFunctions, Planner, Memory, and tools. Setting up an agent project with .NET, Azure OpenAI, and Qdrant. Developing initial tool classes. Mini use case: conversational agent with basic capabilities.
Creating and storing embeddings (text to vector) and integrating Qdrant as a vector store. Building a semantic memory for agent knowledge. Retrieval and prompt context enrichment (live RAG). Implementing tool compositions and dynamic query flows with function calling and Planner.
Agents with complex behavior: Planner, tool selection, and dynamic workflows. Building an interactive interface (e.g., Blazor or Web API). Combining tool functions for data retrieval, decision logic, and response generation. Security, governance, and scaling in the Azure environment. Final demo: your team presents its own agent prototype.