Rabiya Basri PM on 07 Feb 2025 10:55:54
There is a limitation where Q&A index is built based on the first 1,000 model entities (tables and fields). If the data model exceeds this number, the index doesn't hold all the potential model entities, which might decrease Q&A accuracy.
It is not feasible when working with large models. We need help to overcome this limit as Q & A visual doesn't show the search record when index reaching the limit.
Reference document: https://learn.microsoft.com/en-us/power-bi/natural-language/q-and-a-data-sources?source=recommendations