User Inputs - Get Started

Passing Data or Instructions into a Workflow

For almost all workflows, you need some input. These inputs serve different purposes, such as setting instructions, providing actual data to analyze, or asking specific questions. AgenticFlow AI supports a variety of input types, with all requirements taken care of. All you need to do is add the suitable user input component to your workflow.

Creating a Workflow and Adding User Inputs

To get started with creating a workflow and adding user inputs:

1. Navigate to the Workflows Page: Click on + Create Workflow at the top right of the page.

2. Empty State: You will be presented with an empty state where you can:

   • Add knowledge

   • Add inputs

   • Add workflow actions

   • Access some video tutorials

Types of User Inputs

AgenticFlow AI supports various user input types, each fulfilling different purposes:

• Setting and Instruction: For example, specifying the maximum number of items to extract or providing instructions for formatting the output.

• Actual Data to Analyze: For example, uploading a CSV file.

• Questions to Answer: For example, entering a question for the workflow to answer.

Adding User Inputs

1. Add Knowledge: Click on the + Add Knowledge button to provide any information that a language model might not already have.

2. Add Inputs: Click on the + Add Input button to add user inputs. AgenticFlow AI provides a variety of input types, including text inputs, file uploads, and more.

3. Add Workflow Actions: Click on the + Add Action button to add the desired actions to your workflow.

Configuring User Inputs

All user input components accept a name and a description. Descriptions support limited markdown syntax, such as bold, italics, and URLs, to make your inputs more informative and engaging.

Example Configuration:

• Name: my_text_input

• Description: Enter the text to analyze.

Using Variables

Variables are used to reference user inputs in workflows. Variable names are shown in green and can be used with {{}} to access the input values. For example, if you have a text input named my_text_input, you can use the variable name as {{my_text_input}} in your workflow actions.

Presentation of Inputs

• App Page: Inputs are presented as forms where users can enter the required data or instructions.

• API: Inputs are presented as JSON parameters, allowing for programmatic access.

Example

1. Add a Text Input:

   • Click on + Add Input and select Text Input.

   • Configure the input with a name, description, and variable name.

   • Example: Name the input topic, with a description "Enter the topic for analysis."

2. Add an LLM Action:

   • Click on + Add Action and select LLM.

   • Configure the LLM action with the prompt:

     Create an engaging Instagram caption based on the following: "{{topic}}"

3. Run the Workflow:

   • Save the workflow.

   • On the App page, enter a value for the topic input (e.g., "AI is changing the way teams work").

   • Run the workflow and observe the output generated by the LLM.

By following these steps, you can effectively create workflows and add user inputs to enhance the capabilities of your AI-driven automation in AgenticFlow AI.