Conditional Step
A Conditional Step is used when a specific rule or condition needs to be verified, using AI. These are typically complex conditions that require an interpretation where an AI model would provide it more easily in comparison with any other form of condition verification.
For example, checking the Sentiment in a given statement and inferring whether it is a positive sentiment vs a negative sentiment can be achieved easily with the Sentiment Analysis model instead of parsing all the words in a sentence using keywords.
In the above example, using a Sentiment Analysis model to check for the underlying sentiment in a sentence is a conditional step in an AI Automation workflow.
As shown in the Creating a New Data Machine section, to create a Conditional Step, simply drag and drop the Conditional step from the toolbox in the Editor Grid. Once the Conditional Step block is added to the Editor grid, the information section to the right of the screen will present the options to configure the step with your desired functionality.
In order to complete the setup of a Conditional Step, the following information needs to be configured.
Select the inputs required for the execution of the Operational Step. The inputs are dependent on the AI model selected from the prior list input. Review the inputs for each AI model in the Model library to ensure all the required inputs are added.
Note: If the Operational Step being added is the first step in the sequence of the Data Machine workflow, the configuration of inputs is not needed as the inputs are automatically added.
For subsequent steps, whether the step is an Operational or a Conditional step, input parameters need to be configured to ensure the right output from the previous step is captured for the current step.
After the configuration of the Operational Step is complete, another Operational, Conditional or Final Step can be added connecting to the step just added by dragging and dropping from the toolbox.
