Sit in the analyst's chair.
Each case reveals a screening decision stage by stage — input, normalization, candidates, scoring, disposition, investigation. Every person, entity, and list entry is fictional; the reasoning is the real thing being taught.
- 01NAME MATCHINGA clean pass through the filterMost payments pass screening untouched: the filter generates weak candidates, scores them, and discards them below threshold without any human involvement.
- 02FALSE POSITIVESA common surname collides with the listA high name score on a common name is a starting point, not a conclusion — date of birth and geography are what separate the customer from the listed person.
- 03NAME MATCHINGThe alias that was the real nameListed persons rarely transact under their primary list name — a well-maintained alias set is what catches them, and secondary identifiers are what confirm it.
- 04SECONDARY IDENTIFIERSStrong name, wrong person: identifiers decideA near-perfect name score cannot outweigh hard identifiers — date of birth and passport data on the list entry are what let you rule a customer out decisively.
- 05OWNERSHIP & CONTROLNot on the list, still blocked: ownership adds upAn entity owned 50 percent or more in the aggregate by listed persons is treated as blocked even though it appears on no list — the investigation, not the filter, is what finds this.
- 06MESSAGE SCREENINGThe hit hiding in the remittance lineScreening covers every field that can carry a name, and a hit on a non-party in free text raises a question the message alone cannot answer: what role does that entity play in the underlying transaction?