# How to use this accelerator

# Train ML model

Follow these steps to train an ML (machine learning) model.

1

In Workato, create a new recipe.

2

Choose a trigger according to your use case.

3

Add an action and app. Select AIML Accelerator as the application and Train AIML Model as the action.

4

Select a Model Identifier as defined in the AIML | Data Fields lookup table.

5

Configure an if condition to check for the is_success flag. The is_success flag returns true if the preceding action was successful and false if it was unsuccessful.

6

Add the actions for success or failure accordingly. For example, if is_success returns as false, you may decide to send a Slack message or email notification.

7

Click Save.

8

Click Test to run the recipe.

# Stage training data

1

In Workato, create a new recipe.

2

Choose a trigger according to your use case.

3

Add an action and app. Select AIML Accelerator as the application and Train AIML Model as the action.

4

Select a Model Identifier as defined in the AIML | Data Fields lookup table.

5

Configure an if condition to check for the is_success flag. The is_success flag returns true if the preceding action was successful and false if it was unsuccessful.

6

Add the actions for success or failure accordingly. For example, if is_success returns as false, you may decide to send a Slack message or email notification.

7

Click Save.

8

Click Test to run the recipe.

# Get a prediction

Follow these steps to get a prediction.

1

In Workato, create a recipe.

2

Add an action and an app. Select AIML Accelerator as the application and Train AIML Model as the action.

3

Select the Model Identifier as defined in the AIML | Data Fields lookup table.

4

Load the values from the fields in the AIML|Data Fields lookup table into the input fields. Ensure the values match the corresponding Model Identifier. If fields are missing, check the AIML | Data Fields lookup table.

5

Configure an if condition to check for the is_success flag. The is_success flag returns true if the preceding action was successful and false if it was unsuccessful.

6

Add the actions for success or failure accordingly. For example, if is_success returns as false, you may decide to send a Slack message or email notification.

7

Click Save.

8

Click Test to run the recipe.


Last updated: 12/28/2022, 5:28:45 AM