| Creating business functions |
Business functions are only available in decision services that have their language set to
English. |
None |
| Adding the decision logic |
The following features have limited locale support:
- Stream operations are not supported in Chinese.
- String operations are not supported in Chinese, Italian, Portuguese (Brazil), and Japanese.
|
If you require support for these features in currently unsupported languages, submit a
request for enhancement through the appropriate channels. |
| Composing and reusing models |
You cannot reuse task models in decision models. |
None |
| Reusing decision artifacts across decision automations |
You cannot reuse predictive models in other decision automations. Do not publish these models
from the Dependencies tab. |
None |
| When a decision artifact is reused in a decision automation, deleting the original decision
artifact breaks the automation that references it. |
Always audit dependencies before you delete shared decision artifacts to prevent breaking
existing automations. |
| Reusing decision artifacts across decision services and automations |
When you reuse a published decision or predictive model that references data types that are
already used in the target model, duplicate entries appear in the Output type selection dropdown.
This duplication leads to errors, which are displayed in the Error report
tab. |
Remove the duplicated types from the data model used by the target model to resolve the
conflict. |
| Interacting with Decision Designer through an MCP server |
The MCP server supports English only. |
None |
The MCP server does not support the manipulation of:
- Task models
- Predictive models
- Generative AI models
- Business functions in data models
- Function nodes in decision models
|
None |
| Importing industry samples |
When importing the Banking or the Telecom
sample, the import fails and no decisions services are loaded. |
This issue occurs when the Sample Machine Learning Provider is not
available in your tenant. To resolve this issue:
- Create a new decision automation.
- Import the sample again.
The import completes successfully and the decision services are loaded correctly.The
Sample Machine Learning Provider has now been created.
|