This MicroHack scenario walks through the use of Azure OpenAI Services and Semantic Kernel with a focus on the best practices and the design principles. Specifically, this builds up to include working with an existing infrastructure.
This lab is not a full explanation of Semantic Kernel as a technology, please consider the following articles required pre-reading to build foundational knowledge.
Semantic Kernel Docs (read this after completing this lab to take your learning even deeper!)
Prompt Engineering (to help out creating the prompts)
This MicroHack scenario walks through the use of Azure OpenAI Services and Semantic Kernel with a focus on the best practices and the design principles. It will guide you throught the concepts of Generative AI and how to use it in your applications. You'll learn how to use Semantic Kernel to integrate your Azure OpenAI models into your applications, either to generate text or to understand and interpret text.
After completing this MicroHack you will:
- Know how to build a Semantic Kernel application with C# and .NET
- Understand Generative AI concepts such as prompt engineering, embeddings and Retrieval Augmented Generation (RAG)
This MicroHack has a few but important prerequisites
In order to use the MicroHack time most effectively, the following tasks should be completed prior to starting the session.
- Access to Azure OpenAI Service models
- A GitHub account (if you like to use GitHub Codespaces)
- VS Code with DevContainers extension installed (if you want to run it locally)
Use these Settings
Go to Setup to setup your environment. You will find the instructions to setup your environment and the required resources.
Go to Challenge 1 to start the first challenge. You will learn how to initialize the Kernel and how to run Semantic Functions.
Go to Challenge 2 to start the second challenge. You will learn how to generate images with DALL-E 3 and interact with images with GPT-4 Vision.
Go to Challenge 3 to start the third challenge. You will learn how to chat with structured data using Plugins and Planners, that are advanced features of Semantic Kernel that will allow you to create more complex AI applications.
Go to Challenge 4 to start the fourth challenge. You will learn how to chat with your own data using the Semantic Kernel. In this challenge you will be introduced with the concepts of Retrieval Augmented Generation (RAG) and how to use it to chat with your own data.
Go to Challenge 5 to start the fifth challenge. You will learn how to create a Multi-Agent Chatroom using the Semantic Kernel. In this challenge you will be introduced with the concepts of Multi-Agent Chatroom and how to use it to chat with multiple agents.
Congratulations! You finished the Generative AI MicroHack. We hope you had the chance to learn about the how to implement a successful application using Azure OpenAI Services and Semantic Kernel. If you want to give feedback please dont hesitate to open an Issue on the repository or get in touch with one of us directly.
Thank you for investing the time and see you next time!