# Set up a decision model
Use the following sections to create and configure a decision model:
# Create a decision model
Complete the following steps to create a decision model:
Go to Projects > All assets.
Click Create > Decision model.
Enter a Decision model name.
Enter a Decision model name
Use the Location drop-down menu to select the project where you plan to store the decision model.
Click Start building. The decision model builder opens.
Refer to the Configure a decision model section to configure the model.
# Configure a decision model
You can access decision models from Assets > Decision models or from a specific project.
Refer to the following guides to configure a decision model:
Create an input field
# Create an input field
Input fields define the data that enters your model. These fields are available to all nodes, including nodes not directly connected to the Input node. You can also select input fields in the Output node to return them to the calling recipe. This approach helps when the value isn't directly available in the calling recipe, such as when a formula calculates the value at runtime.
Complete the following steps to create an input field:
Select the Input node.
Click Create field if there are no existing fields. Otherwise, click + (Create field).
Create a new field
Enter a Field name.
Input field configuration
Optional. Edit the automatically generated Field label that identifies the field.
Select the field's Data type. Refer to the Available operators documentation to see the condition operators each data type supports.
Optional. Enter a Hint that explains the field. Hints display in decision tables and help users understand the purpose of each field.
Optional. Click the Set as required toggle to make the field mandatory when calling the model from a recipe.
Click Create field.
Delete an input field
# Delete an input field
Complete the following steps to delete an input field:
REMOVE INSTANCES BEFORE DELETING
You can't delete an input field that's in use. You must remove the field from all nodes before you can delete it.
Select the Input node.
Hover over the field you plan to delete and click Delete field (trash icon).
Click Delete field (trash icon)
Click Delete in the confirmation modal.
Create a model field
# Create a model field
Model fields define the data passed between nodes. Model fields are required even in simple single-table models, where they carry table outcomes to the Output node, and become increasingly important in more complex models where they pass outcomes between nodes as intermediate values.
OVERRIDING VALUES
You can reuse a model field as an outcome in multiple nodes. Each node can overwrite the field value.
Complete the following steps to create a model field:
Go to the Model fields sidebar. This menu displays by default when a node isn't selected.
The Model fields sidebar
Click Create field.
Enter a Field name.
Optional. Edit the automatically generated Field label that identifies the field.
Select the field's Data type. Refer to the Available operators documentation to see the condition operators each data type supports.
Optional. Enter a Hint that explains the field. Hints display in decision tables and help users understand the purpose of each field.
Click Create field. You can now add this field to nodes in the model.
Create a Decision table node
# Create a Decision table node
Decision table nodes define conditional business rules. They process data (Condition fields) and output data (Outcome fields) according to business logic you specify.
Complete the following steps to create a Decision table node:
Go to the Toolbar and drag a Decision table node into the model builder.
Click and drag between edges to configure the order of nodes. You can hover over a node connection and click Delete edge to remove it.
Add the node to the model flow
Select the Decision table node.
Optional. Edit the automatically generated table Name.
Decision model node configuration
Optional. Enter a Description for the table.
Use the Condition fields section to select the fields that the table evaluates. You can select fields from the Input node or from the outcome of a directly connected upstream node.
Use the Outcome fields section to select fields to pass from this table to the next node. All Model fields are available, including those already used as conditions in the same table.
Click Configure decision table to open the decision table. You must add at least one condition field and one outcome field before you can open the table.
Refer to the Configure decision table rules documentation to configure decision table evaluation logic.
Configure output fields
# Configure output fields
Output fields define the data that the model returns to the recipe that called it.
Complete the following steps to configure output fields:
Select the Output node.
Click Select fields.
Select which Input and Model fields to return to the recipe that called the model. All input fields are available. You can only select model fields from nodes directly connected to this node.
Click Apply changes.
Create conditional branches
# Create conditional branches
Conditional branches define a set of conditions that control the flow of data in the model using business logic you specify. They allow you to route data to different decision tables based on prior decisions, input fields, or model fields passed from upstream nodes. This makes it possible to handle variations in business logic within a single model, ensuring consistency and ease of maintenance.
FIELDS ARE PASSED UNCHANGED
The condition node passes all received fields unchanged to branches that evaluate to True.
Complete the following steps to add conditional branches to a decision model:
Go to the Toolbar and drag a Conditional branch node into the model builder.
Click and drag between edges to configure the order of nodes. You can hover over a node connection and click Delete edge to remove it.
Add the conditional branch to the model flow
Use the node's drop-down menu to select one of the following options:
- Every matching condition: Runs every branch that matches a condition.
- First matching condition: Runs the first branch that matches a condition.
The conditional branch drop-down menu
Optional. Click + Add condition to add additional conditions.
Each condition you create can contain multiple evaluation criteria. For example:
Each condition you create can contain multiple evaluation criteria.
Select a condition to open its configuration sidebar.
The condition configuration sidebar
Use the Input drop-down menu to select a field. All input fields are available. You can only select model fields from nodes directly connected to this node.
Use the Condition drop-down menu to select an operator. The condition is treated as True if you leave this field blank. Refer to Available operators to see the operators available for each data type.
Enter a Value to compare against the Input using the Condition.
Optional. Click the + (plus) icon to create an additional condition, then select one of the following:
- AND: Inputs are treated as
Trueif they satisfy all conditions. - OR: Inputs are treated as
Trueif they satisfy any condition.
# Model fields
Model fields define the data that moves between nodes in a decision model. Unlike input fields, which receive data from the calling recipe, model fields store intermediate values produced by decision tables and pass them to downstream nodes.
The following describes how data flows through a decision model:
- The calling recipe provides values for input fields.
- Input fields are passed to the decision tables and conditional branches directly connected to the Input node.
- Decision tables and conditional branches evaluate conditions using input fields or model fields from upstream nodes.
- Decision tables write outcomes to model fields and pass them to downstream nodes. Conditional branches pass all received fields unchanged to branches that evaluate to
True. - The Output node returns selected input and model fields to the calling recipe.
Refer to Create a model field to add fields to a model.
# Troubleshooting
Refer to the following sections to troubleshoot issues related to model field sources:
# Missing source
A field becomes invalid when its source is unavailable. This occurs when a connection breaks or an upstream node no longer outputs the field. Restore the source connection to make the field valid again.
An invalid output field
An invalid condition field# Source conflict
A source conflict occurs when multiple decision tables update the same field in a downstream node. The model continues to function, and the downstream node uses values from the decision table that appears first alphabetically. You can remove one of the sources to resolve this conflict.
A source conflict
Alternatively, you can use conditional branches to ensure the source nodes are mutually exclusive. The conflict warning still appears, but only one path carries data at a time, so no conflict occurs at runtime.
Mutually exclusive source nodes
Last updated: 3/31/2026, 6:29:38 PM