Normalize the customer name by removing diacritics, expanding abbreviations, and generating common transliterations before submitting to the screening API
POST the normalized name along with date of birth and country fields to your AML screening vendor's search endpoint
Review the match results array; each potential match contains a match score, the matched list name (SDN, PEP tier, adverse media category), and the matched entity record
Apply a fuzzy match threshold policy to separate strong matches requiring immediate action from weak matches requiring manual review
For PEP matches, classify the PEP tier and determine whether your risk policy requires enhanced due diligence or just additional documentation
Log the full screening result including the request payload, response, match scores, and the disposition decision made by your compliance team
Known gotchas
Name-only screening produces high false positive rates for common names; always include date of birth and nationality when available to improve precision
OFAC SDN list is updated frequently; ensure your vendor syncs list data at least daily and that you understand the data currency SLA in your contract
Adverse media screening requires configuring the relevant risk categories (financial crime, sanctions evasion, fraud, etc.); receiving all adverse media without filtering creates unmanageable noise
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp