From the course: Semantic Kernel in Action: Fundamentals

Unlock the full course today

Join today to access over 24,600 courses taught by industry experts.

Developing prompt functions

Developing prompt functions

From the course: Semantic Kernel in Action: Fundamentals

Start my 1-month free trial Buy for my team

Developing prompt functions

“

- [Instructor] Think of an orchestra, where each instrument adds a unique touch to the symphony. In Semantic Kernel, prompt functions are one of the main types of these musical instruments, adding a custom functionality to our AI applications. Today, we are jumping into the art of crafting these essential elements and see how they are modernized within our AI orchestra. Prompt functions in Semantic Kernel are one of the ways to add functionality to plugins. Essentially, it is a function wrapping a prompt, a request to an LLM, and this is done in a way that it can be used by a software, such an AI orchestrator, which Semantic Kernel is, in our case. These prompt functions allow us to define specific behaviors and responses for our AI models, adding layers of customization and control. We generally do these by adding a prompt and natural language requisite to an AI. Prompts can be formed by using a template, which can include some input variables that determine what to do and its…

Contents