Exception Management Now Supported in Sandbox Accounts
Exception Management is now available in Sandbox accounts, using production-trained models to evaluate workflows safely.
Exception Management is now supported in Sandbox accounts, allowing teams to evaluate workflows using production-quality results without the need for model training or historical data in the Sandbox.
How Does Exception Management Work in Sandbox Accounts?
NetSuite has introduced a feature flag named Exception Management Models, which works alongside the existing Exception Management feature. This separation allows teams to conduct inference without duplicating model training.
To utilize Exception Management (EM) in a Sandbox Account:
- Enable Exception Management Models in your production account, which will initiate model training using production data.
- Then, in the Sandbox account, enable Exception Management to activate the UI for EM.
In the Sandbox environment, EM will run inference using the models trained in production. The system will backfill data for the last two weeks and then continue with hourly updates for newly created or edited transactions.
Important Considerations
- You cannot enable the EM UI feature in both production and Sandbox accounts if Exception Management Models is not enabled in production.
- Additionally, disabling Exception Management Models in production will automatically disable the EM UI feature in both production and linked Sandbox accounts.
- Model training occurs only in production and requires a substantial historical transaction data set for optimal performance—typically, around 18 months' worth.
- Make sure to grant appropriate permissions and role-based visibility since they apply the same way as in production.
Prerequisites
- Ensure that Exception Management AI Models is active in your production account, which requires the Administrator role to activate.
- Production accounts must possess adequate historical transaction data to facilitate model training.
Steps to Enable EM in Sandbox
-
Enable Exception Management AI Models in Production
- Navigate to Setup > Company > Enable Features > Accounting.
- Under Advanced Features, activate Exception Management Models.
- Click Save and confirm if prompted to acknowledge this is for Sandbox testing.
-
Enable Exception Management in the Sandbox
- In your Sandbox account, go to Setup > Company > Enable Features > Accounting.
- Under Advanced Features, turn on Exception Management to activate the UI.
- Click Save.
-
Review EM Transactions
- Access the EM dashboard in Sandbox to analyze exceptions using transactions trained in the production environment.
Availability and Compatibility
This feature change is backward compatible, meaning any existing account with EM previously enabled will have both session flags activated by default. If you have queries or do not see these options, consult your NetSuite Account Manager for assistance. Exception Management remains a limited-release feature contingent on data volume and model readiness.
Key Takeaways
- Exception Management is now functional in Sandbox accounts, leveraging production data for inference.
- Model training occurs solely in production, reducing overhead for Sandbox environments.
- Feature flags govern model training and UI access, emphasizing a clean separation.
- Standard prerequisites apply, ensuring proper permissions are set up.
Source: This article is based on Oracle's official NetSuite documentation.
Frequently Asked Questions (4)
Do I need to activate a feature flag in production before enabling Exception Management in Sandbox?
What are the prerequisites for enabling Exception Management in a Sandbox account?
Can model training occur in Sandbox accounts?
What happens if I disable Exception Management Models in production?
Was this article helpful?