Develop AI Agents using Azure OpenAI and Semantic kernel

This course provides a comprehensive guide to building intelligent AI agents using Azure OpenAI and the Semantic Kernel SDK. Participants will learn to develop applications that automate tasks and perform natural language processing by integrating large language models with custom plugins and planners. The curriculum emphasizes practical skills in prompt engineering, plugin development, and responsible AI implementation, ensuring learners can create effective and ethical AI solutions.

By the end of this course, learners will be able to:

• Understand the purpose and capabilities of the Semantic Kernel SDK
• Develop and optimize prompts for AI interactions
• Create and integrate plugins to extend AI agent functionalities
• Implement native functions to provide skills to AI agents
• Combine prompts and functions to enhance AI responses
• Utilize planners to automate function calls and optimize workflows
• Develop a proof-of-concept AI application using the Semantic Kernel SDK

Participants should have:

• Experience programming in C#
• Visual Studio Code IDE installed
• Familiarity with Azure and the Azure portal
• Access to Azure OpenAI Services

Target audience

This course is suitable for:

• AI engineers, solution architects, and software developers
• Technical professionals interested in generative AI applications
• Organisations aiming to improve service automation and internal productivity with custom AI solutions

Module 1: Build your kernel

• Introduction to Semantic Kernel
• Understanding the role of the kernel in AI applications
• Setting up the development environment
• Creating a kernel object
• Exercise: Build a kernel object

Module 2: Run prompts with Semantic Kernel

• Basics of prompt engineering
• Using prompt templates
• Implementing Handlebars prompt templates
• Storing and managing chat history
• Exercise: Use Semantic Kernel prompt templates

Module 3: Create Semantic Kernel plugins

• Understanding native plugins
• Developing native functions
• Configuring function choices
• Exercise: Create native functions

Module 4: Apply filters on functions

• Combining prompts and functions
• Implementing prompt injections
• Applying trust filters
• Filtering invoked functions
• Exercise: Add a filter service to the kernel

Module 5: Guided project - Create an AI travel agent

• Preparing for the guided project
• Developing a currency converter
• Creating a travel itinerary
• Providing context cues
• Exercise: Develop an AI travel assistant

Exams and assessments

There are no formal exams included in this course. Learners will complete interactive labs, guided exercises, and scenario-based tasks to reinforce understanding and assess their progress.

Hands-on learning

This course includes:

• Guided labs on kernel creation, prompt development, and plugin integration
• Practical exercises for implementing filters and planners
• Simulated real-world scenarios for developing AI applications
• Instructor feedback and collaborative learning activities