1.2 Customer Screening Workflow

The figure illustrates the Customer Screening workflow.

Figure 1-2 Customer Screening Workflow


Customer Screening Workflow

Customer screening workflow orchestrates the sanctions screening of customers against internal and external watchlists, ensuring.

This workflow ensures:
  • Automated, scalable screening across multiple tenants/clients.
  • Consistent data quality and processing through rigorous controls.
  • Strict role separation and robust security to prevent data leakage.
  • Thorough investigative and compliance traceability. .

Here’s a step-by-step explanation:

Data Ingestion
  • The data ingestion process helps to collect and compartmentalize data for each tenant/client to uphold strict data separation and confidentiality.
  • Data Sources:
    • Customer Lists: Core customer/client records subject to scrutiny.
    • Party Lists: Linked individuals/organizations of interest (e.g., beneficial owners, related vendors).
    • Transactions: Counterparties in a transaction.
  • Watchlists:
    • External: Sanctions and PEP lists (e.g., OFAC, EU, UN, government sources) are securely downloaded via HTTPS. FinCEN names are uploaded by users in the application upon downloading the FinCEN list from a secure web-based interface.
    • Internal: Proprietary/private lists managed by the organization that can be managed through UI and API.
  • Ingestion Mechanism:
    • Users securely upload CSV data to the Oracle Cloud Infrastructure (OCI) Object Store, partitioned by tenant to ensure data privacy and access control.

Staging, Business & Data Quality Process

This process allows users to prepare data for reliable screening while enforcing quality and business rules.

  • Data Pipeline:
    • Cleansing: Standardizes names, dates, and removes inconsistencies (e.g., whitespace, encoding errors).
    • Validation: Ensures data is complete, accurate, and meets business rule requirements (e.g., mandatory fields, data types).
    • Enrichment (optional): Additional context or identifiers can be appended.
  • Categories Maintained:
    • Party/Customer Data: Structured information about individuals or organizations.
    • External entity data: Structured information about Transaction counterparties
    • Watchlists: External and internal, consolidated for screening.
Index Loading & Search Preparation
  • Data Cleaning & Validation: Verifies only high-quality, valid, and standardized data moves forward
  • Index Loader & REST Service:
    • Cleaned data is relayed via REST APIs to an intermediate 'Index Loader.'
    • This decouples upstream processing from the search engine, allowing for asynchronous and extensible data flows.
  • Logstash Integration: Leverages Logstash for advanced data transformations, enriching, and loading into OpenSearch.Pipeline: HTTP Input → Filter/Transform → Output/OpenSearch Index
  • OpenSearch Index:
    • Enables rapid searching, dynamic filtering, and analytics.
    • Designed for scalability, privacy, and auditability.
Screening & Matching Engine
  • Modes of Screening:
    • Bulk (Batch) Matching: For scheduled/periodic, high-volume processing.
    • Real-time (Single) Matching: For “instant” checks—via API and UI—for onboarding or transactional triggers.
  • Matching Rule Management:
    • Role-based secured environment for configuring matching parameters (algorithms, thresholds, fuzzy logic).
    • Change management and auditing for configuration changes.
  • Screening Execution:
    • Matching engine queries OpenSearch for candidate matches on legal/known names, aliases, and other relevant identifiers.
    • Results surfaced through both REST API and user interface.
Alerting, Case Management, and Investigation
  • Alert Routing: Positive/Partial matches (potential alerts) are automatically sent through REST API to the case management system (e.g., iHub).
  • Case Creation: Cases triggered for individuals, entities, vessels, and external entities on:
    • Sanction matches
    • PEPs
    • Inclusion on adverse media/proprietary lists
    • Other exclusion/warning triggers
    • Prohibited country lists
    • FinCEN names.
  • Investigation:
    • Investigators securely access the case management environment with strict logging and access controls.
    • Full audit trails maintained for all actions: assignment, escalation, investigation, and closure.
    • Cases can be worked, documented, and escalated as needed according to defined procedures.
Security, Access Control & Audit
  • Tenant Isolation: Logical separation of data for each client/unit to guarantee confidentiality.
  • Secure Transmission/Storage: HTTPS for all data flows; encryption at rest and in transit for sensitive data.
  • Role-based Access: Least-privilege principle is applied across ingestion, configuration, investigation, and administration.
  • Audit Logs: Detailed logs of key activities (ingestion, rule changes, investigations) for compliance and forensic review.

Audit

Comprehensive audit history maintained for:
  • Cases created in the Investigation hub, for real-time and batch screening
  • Private watchlist management
  • Batch execution and screening
  • Real-time screening