Match thresholds are about to freeze for production. What testing step should precede that decision?
Select an answer to reveal the explanation.
Short Explanation
Before you lock the dials, have business folks eyeball real match samples. Unit tests on fake strings won’t catch your city’s mess.
Full Explanation
Before match thresholds freeze for production, business users should review representative match samples so precision and recall are validated. Civic stewards—not only engineers—own identity risk acceptance. Freezing thresholds with no business sample review to save calendar time fails quality gates. Letting only the database vendor approve merges without domain experts fails business ownership. Skipping UAT because synthetic random strings looked clean in a unit test fails realism. Exam caveat: business sample validation precedes threshold freeze. Operational check: prepare stratified samples (easy matches, near-misses, known non-matches), facilitate steward review, record precision/recall, then freeze with sign-off. Store signed sample results in the municipal UAT evidence pack. Keep verification evidence with the municipal runbook so audits can prove the chosen control was applied as designed and that rejected alternatives would have left city services exposed to the failure modes described above. Keep verification evidence with the municipal runbook so audits can prove the chosen control was applied as designed and that rejected alternatives would have left city services exposed to the failure modes described above.