AZ-2005: Develop Generative AI solutions using Azure OpenAI and the Semantic Kernel SDK

Course Description

Harness the power of Azure OpenAI and Semantic Kernel SDK to build intelligent, generative AI applications. This one-day, instructor-led course teaches developers to construct AI agents capable of executing sophisticated tasks using natural language prompts, native functions, and planning workflows. You'll develop skills in building a robust kernel, creating plugins, combining prompts and code, and orchestrating automations—all with a focus on real-world implementation.

Target Audience

This course is perfect for:

• AI Developers, Software Engineers, and Cloud Developers building intelligent applications using Azure OpenAI and Semantic Kernel

• Professionals proficient in C#, familiar with Visual Studio Code and Azure portal

Prerequisites:

• Strong C# programming experience

• Azure OpenAI Service access

• Familiarity with Azure development environments 

Course Outline

Module 1: Build Your Semantic Kernel

• Learn the fundamentals of the Semantic Kernel SDK

• Initialize a kernel that connects C# code with large language models

• Explore best practices for prompt engineering and kernel architecture 

Module 2: Run Prompts with Semantic Kernel Plugins

• Use and configure built-in prompt plugins

• Apply prompt templates and store chat context

• Enhance applications with reusable prompt structures 

Module 3: Create Native Function Plugins

• Develop native functions (code-based plugin skills)

• Combine functions with prompts to extend AI capabilities

• Embed custom business logic into AI workflows

Module 4: Apply Filters and Secure Function Calls

• Implement trust filters to manage function invocation

• Combine filters and prompt logic for secure, robust operations

• Learn validation and safety checks in AI agent workflows

Module 5: Use Intelligent Planners

• Define and configure planners within the Semantic Kernel

• Automate function orchestration using planning logic

• Optimize agent automation with prompt-function chaining

Module 6: Guided Project—Create an AI Travel Agent

• Implement an end-to-end proof-of-concept AI agent using kernel, prompts, functions, and planners

• Build capabilities like itinerary planning, currency conversion, and decision logic

• Tie together Azure OpenAI, Semantic Kernel, and code-based plugins

HandsOn Experience

Approximately 40–50% of the class is hands-on, with practical coding sessions and live demos designed to reinforce skills in building generative AI solutions with Azure OpenAI and Semantic Kernel.

Skills You’ll Gain

By the end of AZ2005, you will be able to:

• Initialize and manage a Semantic Kernel for AI development

• Create and run both prompt-based and native function plugins

• Apply trust filters and secure function calls in your agents

• Use planners to automate workflows and chain intelligent operations

• Construct complete AI agents combining prompts, code, and planning for real-world scenarios