Currently, the OACS AI Assistant in Gen2 Platform operates by querying the entire Subject Area, which often results in unstructured outputs and increases the risk of hallucinations or inaccurate responses. This poses a challenge for business users who rely on data accuracy and may lead to a high volume of queries regarding data quality and correctness.

Proposed Enhancement:
We propose enabling the AI Assistant to work directly on the output of structured reports—such as ad-hoc reports or dashboard reports—created by business or technology teams. Instead of querying the entire Subject Area, the Assistant would analyze the already curated and validated data presented in these reports to generate insights.
This approach would:

Reduce the risk of hallucinations and incorrect answers.
Improve trust and adoption among business users.
Provide meaningful, contextual insights based on verified data.
Enhance the value of existing reports by enabling deeper exploration without compromising data integrity.

Use Case Example:
A business user views a dashboard showing quarterly sales performance. By interacting with the AI Assistant, they could ask:

“What are the top 3 factors contributing to the dip in Q2 sales?”
“Can you summarize regional performance trends from this report?”

Request:
Please consider enabling this feature in the OACS AI Assistant to allow insight generation based on structured report outputs rather than querying the entire Subject Area.