We provide comprehensive validations of key BSA/AML model components, including:
Transaction Monitoring Systems
We assess the model's ability to detect potentially suspicious activity by reviewing:
Scenario logic and typology coverage (e.g., structuring, layering, rapid movement of funds)
Thresholds, parameters, and tuning
Alert generation rates, effectiveness, and false positive metrics
Coverage analysis to ensure all relevant products, channels, and customer types are monitored
Data mapping and ingestion from core and ancillary systems
Customer Risk Rating Models
We review methodologies used to assign initial and ongoing risk ratings to customers, including:
Risk factor weighting (e.g., geography, occupation, entity type, product usage)
Scoring logic, segmentation, and overrides
Consistency across onboarding and periodic reviews
Governance over changes to rating models and risk classifications
OFAC / Sanctions Screening Systems
We assess the effectiveness of real-time and batch screening models, including:
Matching logic (fuzzy, phonetic, exact)
List management and updates (OFAC, custom lists)
Hit review workflows
Model tuning and effectiveness of name-matching algorithms
Data Integrity & Mapping
We perform a detailed review of source data quality, field mapping, completeness, and transformation logic. This ensures AML models are relying on accurate, timely, and consistent information across transaction types, accounts, and customer profiles.
Model Governance and Documentation
We evaluate the AML model lifecycle management process, including:
Model documentation and change control
Threshold tuning and scenario governance
Periodic backtesting and effectiveness reviews
Integration into the broader Model Risk Management (MRM) framework
Vendor & Platform Expertise
We validate AML models across a wide range of platforms, including, but not limited to:
Actimize
Verafin
Patriot Officer
BAM+
Fiserv's FCRM
Yellow Hammer
In-house SQL or rules-based systems
