AI

AI actions let you integrate artificial intelligence into your workflows. At their core is the LLM (Large Language Model) action, which can understand text, generate content, analyze data, and help with a wide range of tasks.

How it works

When you add a Prompt action to your workflow, you're creating an interaction with an AI model. You provide instructions (called prompts) and any necessary data, and the model generates appropriate responses. These responses can be anything from simple text to structured data, depending on your needs.

Every Prompt action consists of three main components:

  • The prompt that tells the model what to do

  • The model configuration that controls how it behaves

  • The output settings that determine what you get back

Key capabilities

The Prompt action can handle various tasks:

  • Generate text and content

  • Analyze documents and data

  • Extract information

  • Answer questions

  • Translate languages

  • Help with decision-making

You can enhance these capabilities by engineering prompts to get better results, fine-tuning model settings, structuring output formats, and comparing various prompt and model configurations.

Getting started

To use AI in your workflow:

  1. Add an LLM action where you need AI capabilities

  2. Choose a model appropriate for your task

  3. Write or generate a prompt that describes what you need

  4. Configure how you want the model to respond

  5. Test and refine until you get the desired results

Continue reading:

Using the Prompt action →

Explore prompt engineering →

Understand model configuration →

Last updated