Does this find better tests?
What we know and what we have not measured.
On this page
Does this find better tests?
We have not yet measured whether Qloo-guided discovery finds more useful, repeatable cases than ordinary test generation.
What we know so far
Real runs found restaurant pairs whose order changed between profiles. Another valid run found no useful contrast. The system can save and test those cases, but that alone does not prove it finds better tests or improves diner satisfaction.