Hello,
We are encountering a critical issue with Amazon Q in QuickSight Scenarios.
Context
-
We implemented multiple scenarios in a production QuickSight account for a client
-
These scenarios were working correctly and were actively used
-
No changes were made to:
-
Datasets
-
SPICE ingestion
-
Topic configuration
-
Scenario logic
-
Issue
Recently, all scenarios in the production account started behaving incorrectly:
-
Steps do not execute
-
In some cases, the scenario stops before even completing Step 1
-
No error messages are shown
Key Observation
-
The same scenarios (or equivalent logic) work correctly in a sandbox account
-
The issue affects all scenarios, not just one
What we ruled out
-
SPICE capacity → sufficient
-
Dataset integrity → validated
-
Prompt complexity → even simple steps fail
Question
Has anyone experienced:
-
All scenarios failing or not executing in a specific account?
-
Account-level issues with Q Scenarios?
-
Possible backend changes or regressions affecting execution?
At this point, this appears to be an account-level or service-side issue rather than a modeling problem.
Any insights would be greatly appreciated.



