I want to set a clear threshold (like 0.85) to automatically flag likely duplicates for review.

This feature lets you define a clear similarity threshold (for example, 0.85) that automatically flags items as likely duplicates. It helps reviewers focus on the highest-confidence matches, reducing manual scanning and improving consistency in duplicate detection.

This feature allows you to set a numeric similarity threshold that determines when a pair of items should be automatically flagged as likely duplicates for human review. You choose the threshold value (such as 0.85) based on how strict you want duplicate flagging to be. When similarity meets or exceeds the threshold, the system marks the items for review instead of requiring users to manually search and compare everything. This creates a consistent, repeatable rule for what counts as “likely duplicate,” which can reduce disagreement across reviewers. It also helps teams prioritize review effort by surfacing only the most relevant candidates. A higher threshold typically results in fewer flags with higher confidence, while a lower threshold may surface more potential duplicates for broader review. This is useful in workflows where duplicate records, submissions, tickets, or entities can create reporting noise or operational inefficiency. It supports quality control by ensuring potential duplicates are identified early and routed to a review step. The outcome is a more streamlined review queue and improved data cleanliness over time.