HAWK:AI Transaction Monitoring & Watchlist Screening
Hawk AI is a software platform that combines AI with traditional rule-based approaches to monitor financial transactions in real-time, delivering next-generation anti-money laundering compliance for financial institutions. The solution offers classic rule-based models, which are enhanced by false-positive-reduction features based on machine learning models that learn from the investigator’s own decisions through our case manager. Hawk AI makes use of an unsupervised machine learning model, Anomaly Detection, to identify new patterns of crime through insights from the overarching nature of the platform spanning multiple financial institutions. The platform provides full transparency of machine decisions to deliver the necessary clarity for regulators that require “explainable AI”, as well as instill trust in the machine's decisions. Using Artificial Intelligence to maximize automation, Hawk AI delivers a significant cost benefit through more than 70% reduction of required resources and better crime detection. The result: a more efficient and effective way of fighting financial crime.
The system consists of
(a) the Surveillance Engine,
(b) Case Management,
(c) Configuration Management User Interface.
a) Surveillance Engine
The real-time, rule-based detection of suspicion is integrating the four functional modules:
1. AML Transaction Monitoring
3. Watch list screening (Customer and Payment screening)
4. Customer Risk Rating
5. False Positive reduction
6. Anomaly detection
To augment rule-based functionality, we provide two modules based on comprehensive, state-of-the-art supervised and unsupervised Machine Learning Models, which also operate in Real-time:
False Positive Reduction leads to well-documented, suppressed alerts. The decisions are based on learning effects from the history of case handling by your operators as well as from ongoing operations.
Anomaly Detection allows for the detection of risks previously not detected in the pool of all monitored transactions. This is accomplished by leveraging ML models to form a view of “normal” and “anomalous” behavior within individual accounts and account segments.
(b) Case Management
Alerted cases are displayed for manual investigation in the HAWK:AI Case Manager. The Case Manager is unifying all alerts triggered by the above Surveillance Engine and allows efficient work-out and/or escalation of cases or groups of cases by providing a 360-degree view of each case within the tool.
(c) Configuration Manager
The configuration of rule parameters and Machine Leaning functionality can be performed through the web interface (drag-and-click) in a user-friendly way without the need for programming knowledge. Adding/combining rules and respective parameters is possible with the option to apply different rules depending on e.g., low/mid/high-risk customer risk, or product groups. No code configuration of the rules and patterns applied in the surveillance engine and easy-to-understand the configuration of the Case management
1. Cost reduction – with 70-90% fewer false positives AML/compliance officers waste less time on ‘busy work’ and can focus on compliance tasks that add more business value.
2. Combating crime – AI-powered anomaly detection helps financial institutions fight financial crime by identifying and flagging new threats before they become widespread
2. Explainable audit trails – Our transparent systems and explainable AI (patent pending) provides clear audit trails, ensuring all stakeholders (from frontline officers to auditors, supervising authorities, partners, and regulators)."