Anti-Money Laundering Suite (AMLS)
An end-to-end machine learning-powered AML/CFT software application
AML Compliance Is a Tough Task Amid Rising Complexities
Rapid growth in wire activity is making it difficult to monitor and detect suspicious transactions and names effectively and efficiently, while regulators are enforcing a higher level of scrutiny in the money laundering space. A few noteworthy figures:
- Globally, less than 1% of illicit transfers are being brought to light and the estimated amount of money laundered per year is 2-5% of global GDP, or US$800 billion-US$2 trillion.
- Regulatory fines imposed on them for compliance lapses stood at over US$300 billion since 2008, with US alone accounting for ~15% of the total.
- Currently, banks handle around 150 AML alerts on a daily basis. This is expected to surge due to the growing complexity of compliance regulations. More than 300 million pages of regulatory documents will be published by 2020.
Today, banks mostly rely on manual efforts for their AML compliance but adding more human resources is not a viable long-term solution. Rules-based applications that produce 95% false alerts, creating huge backlogs and massive ageing of alerts, cannot be sustainable either. With legacy systems, the costs of managing the process to read, analyze, and implement the changes in operations will only increase over time.
Tookitaki AMLS: Adding Power to Your Compliance Lens
- Reduces false alerts by at least 40% across monitoring and screening programs.
- Reduces risk by improving the detection of true suspicious cases (SAR/STR) by 5%.
- Prioritizes alerts and brings more efficiency in the alerts disposition process.
- Maintains high detection coverage with growing data and changing regulations through continuous automatic learning.
- Explains complex machine learning models and outcomes in an easy-to-comprehend manner.
- Easy integration to the FI’s existing and future up/downstream systems through built-in connectors and REST API interfaces.
AMLS Modules: Ensures maximum coverage and high alerts yield
Transaction Monitoring: We provide secondary scoring of the transaction monitoring process. The approach uses our proprietary semi-supervised technique which is based on a combination of multi-dimensional unsupervised techniques, network analysis and supervised learning to detect complex money laundering structuring mechanisms. The module complements the current rules-based system by triaging alerts (buckets alerts into L1, L2 and L3 cases) and detecting ‘unknown unknowns’ or true suspicious cases missed by the primary alerts generation unit.
Screening: We provide secondary scoring of the screening process. The approach uses our proprietary multilayered supervised techniques which combine approaches in improved matching techniques (handles typos, spelling errors, titles, prefix/suffix, etc.) and detailed analysis of secondary information obtained either from internal sources or externally available sources. The output prioritizes/triages hits across individual and corporate names into L1, L2 and L3 buckets.