AML Investigation in Action: Pandora Papers Unveiled

Challenge

The institution faced several obstacles that hindered their anti-money laundering intelligence analysis efforts:

• Regulatory Deadline: A fast and accurate response was essential to meet the strict timelines set by regulators.
• Data Complexity: The team had to normalize and match unstructured external data, in this case the Pandora Papers, with internal records. This required overcoming linguistic variations, incomplete data, and dispersed information across siloed sources.
• Limited Investigative Agility: Tools like Quantexa, while suitable for enterprise-wide management, lacked capabilities for sandbox environments and seamless integration of new data sources—key features needed for rapid AML investigation.
• Technical Barriers: Analysts struggled with querying across multiple databases, each with different logic and schema, leading to resource-intensive operations and frequent query failures.
• Reproducibility: Ensuring workflows and queries were auditable and reproducible for future regulatory follow-ups was critical to maintaining compliance and operational accountability.

Solution

DataWalk met the need, and showed how it could serve as a robust framework for anti-money laundering intelligence analysis and streamlined AML investigations through the following features:

• Streamlined Data Integration: DataWalk supported both batch-mode and API-based imports while enabling analysts to manually import unstructured datasets, like the Pandora Papers, into a secure sandbox environment. This hybrid approach allowed rapid setup without disrupting enterprise workflows.
• Entity Resolution and Analysis: DataWalk’s entity resolution engine linked entities across documents, handling name translations and multi-language fonts effectively. Analysts could perform high-level entity resolution within a sandbox, ensuring flexibility in anti-money laundering investigations without affecting the master analytical model.
• No-Code Knowledge Graph and Ad-Hoc Queries: Analysts leveraged DataWalk’s visual, no-code query interface to perform complex AML intelligence analysis across multiple databases. These tools eliminated the need for proficiency in query languages like SQL or Cypher, reducing errors and increasing operational efficiency.
• Collaborative Investigation Models: Analysts built reusable meta-models, dynamically integrating data from KYC, CRM, transactional records, and external repositories, enhancing collaboration and flexibility in AML investigations.
• Reproducibility: DataWalk’s transparent workflows ensured all steps were recorded, reproducible, and modifiable. This ensured compliance with regulatory standards and audit readiness, even months after the initial inquiry.

Figure 1: Knowledge Graph integrating customer data with external 'ad-hoc' datasets, such as the Pandora Papers Solution

Results

DataWalk software was shown to deliver  transformative outcomes in anti-money laundering intelligence analysis:

• Accelerated Response Time: Regulatory inquiries could be resolved in days rather than months.
• Enhanced Investigative Flexibility: Analysts can dynamically test hypotheses and integrate new data without impacting existing analytical models.
• Transparency and Accountability: Fully reproducible workflows ensured compliance with regulatory standards and audit requirements.
• Scalable Adoption: The solution’s Minimum Viable Product became a blueprint for global units within the institution, establishing a scalable framework for combating financial crimes.

This case study illustrates how DataWalk enabled the financial institution to meet stringent regulatory requirements while establishing a next-generation platform for AML investigations. The platform’s powerful anti-money laundering intelligence analysis capabilities transformed the institution’s approach to financial crime investigations, paving the way for scalable, compliant, and efficient operations.