SELECT NINJA PROJECTS

AML MODEL DEVELOPMENT

FINANCIAL CRIMES ENFORCEMENT NETWORK

Ninja Analytics currently supports FinCEN’s Advanced Analytics program through the provision of data scientists who develop statistical and text analytic models for FinCEN.  Over the past three years, Ninja’s analysts have worked hand-in-hand with FinCEN’s analytic cadre to: (1) develop and operationalize new business rules for their SAS Fraud Framework implementation; (2) develop predictive models focused on the extraction of entities and the conceptual categorization of SAR narratives; (3) stand up SAS Social Network Analysis tools to allow investigators and analysts to “walk the graph” in real time; and (4) develop predictive models for surfacing suspicious MSBs and check cashers, among other entities.

 

The team has also been deeply involved in the production of many different strategic reporting initiatives, including analyses of Credit Unions, Money Service Businesses, Check Cashers, and other Intelligence products.  Ninja has provided advanced analytic capabilities in support of multiple legal investigations (e.g., Southern District of NY, State of Florida, Office of the Comptroller of the Currency) focused on money service businesses and third-party payment processors.  Other analytical projects included:  Interpol matching, country benchmarking, unregistered Money Service Businesses, Mexican SAR prioritization, entity resolution, and projects in support of a Local SAR Review Team and the Commodities, Futures and Trading Commission (CFTC).

TOP 5 FINANCIAL INSTITUTION

Ninja Analytics is currently supporting model development and enhancement efforts for a Top 5 U.S. financial institution.  Model development has focused on two primary domains: (1) Indian Banking Association required scenarios; and (2) trade-based money laundering.  In addition, the team is working to trouble-shoot and tune under-performing models in the Bank’s broker-dealer space.  The team’s solutions are being implemented on two platforms:  (1) Oracle’s Financial Crime and Compliance Management (OFCCM) platform (aka as Mantas); and (2) SAS Grid / Teradata platform.

TOP 10 FINANCIAL INSTITUTION

One of Ninja's senior analysts served as the primary model developer for a large financial institution tasked with creating a model to surface suspicious Money Service Business (MSB) patterns embedded in approximately 100 million transactions over a five-year period.  The model developer worked hand-in-hand with the advisory and investigative teams to ensure that the strategies were closely aligned with investigative needs.  The modeling team also worked closely with the bank’s model risk office and an outside validator to ensure model validity.

LEGAL INVESTIGATIONS

FINANCIAL CRIMES ENFORCEMENT NETWORK

Ninja Analytics has provided advanced analytic capabilities to both large financial institutions and the Financial Crimes Enforcement Network (FinCEN) to support multiple legal investigations (e.g., Southern District of NY, State of Florida, Office of the Comptroller of the Currency).  Analytic capabilities included the development of statistical “models” to surface activities and entities of interest and also the provisioning of tools (conceptualizing and instantiating data) for use by investigators.

TECHNOLOGY INTEGRATION

FINANCIAL CRIMES ENFORCEMENT NETWORK

At FINCEN, Ninja analysts created multiple statistical models to identify illicit check cashers specific to a multiple geographic areas.  The team organized the data output by the model via the SAS Fraud Framework and the SAS Social Network Analysis (SNA) tool.  Ninja Analysts conceptualized and instantiated these tools to provide analysts with the capability to “walk the graph” pulling back additional data on command.  Analytical products were also integrated with the desktop version of IBM i2 Analyst Notebook product to facilitate sharing entities and linkages with law enforcement.

MODEL VALIDATION

TOP 5 FINANCIAL INSTITUTION

Ninja evaluated the operation of a large financial institution's scenario validation framework and team interactions.  A number of recommendations emerged from the evaluation.  Ninja recommended:

  • Alignment of the framework to model risk management standards.

  • Clarification of responsibilities and expectations of developers and validators.

  • Establishment of scenario risk ratings and alignment to the Bank's Money Laundering Risk Assessment.

  • Specific methods to track validation responses and completion, and the alignment of priorities to MLRA and policy.

  • Establishment of annual review process and the incorporation of change logs and model failure logs into documentation.

© 2020 by Ninja Analytics, Inc. -- All Rights Reserved