1. ZOLOZ SMART AML

ZOLOZ SMART AML Solution is a comprehensive risk-based, financial-grade AML solution featuring a range of services for your compliance needs, including watchlist screening, transaction monitoring and customer risk rating.By combining our cutting-edge AI technology with our regulator-approved compliance models, we are able to support you throughout the whole cycle of anti-money laundering and help you stay ahead of the ever-changing world of financial crime.

Case Management is integrated into screening, customer risk rating and transaction monitoring.

1.1. Screening

1.1.1. Overall Introduction

Screening is for enterprise to manage economic sanction compliance risk. Through screening customers against watch list/name list with screening engine, it delivers screening matching results and provide corresponding platform for further investigation to prevent assist you in compliance with the local laws and regulations and maintain the confidence, trust and loyalty of its customers, partners and workforce. A good screening solution means you must successfully screen transactions and customers against multiple sanctions lists, identify politically exposed persons (PEPs), and avoid transacting with government sanctioned countries,organizations, and individuals or dealing with embargoed goods.Screening system is for enterprise to manage economic sanction compliance risk. Through screening customers against watch list/name list with screening engine, it delivers screening matching results and provide corresponding platform for further investigation to prevent assist you in compliance with the local laws and regulations and maintain the confidence, trust and loyalty of its customers, partners and workforce. AML screening system employs a score-based screening engine and provides a set of functionalities.

1.1.2. Screen Engine

The core screening engine of the Product employs a set of proprietary text-processing rules and algorithms to match input transaction and customer data against internal and external watchlists. Alerts / hits will be triggered if the input data have matching scores exceeding a pre-defined threshold. These hits will then be handed over to alert investigators to conduct manual review on the case investigation platform (embedded in the Product).

Screening system employs a score-based screening engine, which specifically incorporates a set of text-processing rules and fuzzy matching algorithms to screen input data of various formats against watchlist information. A matching score in the range of 0 to 100 will be returned for each screened record, indicating the level of similarities between the screened record and the selected watchlists. For example, an exact match would generate a matching score of 100. Only the matching results with the score over the pre-defined threshold and rules will be output as hit for subsequent alert review process.

1.1.3. System Function

1.1.3.1. Realtime Screening

The following diagram is the real-time screening workflow within the system.

1.1.3.2. Batch Screening

The following diagram is the batch screening workflow within the system.

1.1.3.3. Dashboard

The Product contains a real-time dashboard to summarize both input and alert data at different dimensions and allow you to monitor key data features. The current dashboard includes the following metrics:

• Key performance indicators for the current day, including the alert volume, match rate, alert review completion rate.

• Screening and Alert review indicators over the past 7 /14 /30 days, including screened volume and alert generated volume.

• The business / Scenario that are being screened over the past 7 /14 / 30 days, including the total number of scenarios and the underlying percentage.

1.1.3.4. Alert Manager

The Alert Manager supports the organizational compliance process for internal review. It is a web-based application used by analysts, compliance professionals, managers, administrators and other principals involved in the alert investigation process.

Alert Manager Service has two main pages：

· Task List Page

· Review Detail Page

Next, this chapter will introduce the Alert Manager and Work-flow according to staff's perspective.

1.2. Transaction Monitoring

Monitoring transactions for suspicious activity is a key element of any AML program. Financial institutions should, at a minimum, acquire and maintain systems adequate to generate accurate, timely, and complete information sufficient to fulfill your obligations to detect and report suspicious activity.

Our AML transaction monitoring module, featuring with flexible function capability and robust infrastructure, is under same purpose to reflect your rational risk assessment and support you mitigate risks more efficiently.

1.2.1. Overview

In general, the transaction monitoring system screens transactions and customer information on a batch basis, generates alerts for investigation, and integrates with a case management tool that allows for alert management and reporting functions.

The high level system diagram is as following:

The following diagram is the Transaction Monitoring workflow within the system.

1.2.2. Dashboard

The dashboard summarizes various metrics and can be used to monitor the money laundering risk at enterprise level and investigation progress.

1.2.3. Model Manager

“Model” in model manager is where the users configure rules to monitor transactions, there are usually two types of models: Transaction Monitoring Model and Result Integration Model.

The Transaction Monitoring Model, as a collection of multiple rules, runs automatically. Users can add rules with different running frequencies into the model, and the edit function is also provided in the model.

The Alert Integration Model has only one rule, aim at integrating transactions which hit by rules in the Transaction Monitoring Model, the results of the integration will be pushed to another module as alerts for reviewing.

1.3. Customer Risk Rating

1.3.1. Overview

Risk Assessment is a module to determine the risk level of the customer, and to create EDD tasks based on pre-set models.

1.3.2. Realtime Customer Risk Rating

The following diagram is the real-time screening workflow within the system.

1.3.3. Periodic Customer Risk Rating

The following diagram is the periodic screening workflow within the system.

1.3.4. Dashboard

CRR Dashboard is used for monitoring the risk class distribution and high risk trend of CRR model.

1.3.5. Model Management

Model is an important part of The Risk Assessment, where the users set up rules to rate customer risk level.

There are usually two types of execution mode to meet different requirement: batch mode and real-time mode. And no matter what kind of execution mode it is, a model can only rate a kind of subject (Individual or Entity), so it's necessary to set up different models for Individual and Entity.

You can find how to create an periodic customer risk rating batch model in 2.2.1, how to set up rules in 2.2.2 Rule Manager, how to create and publish a real-time model in 3.

Before that, you should have already created batch factors or real-time variables. Since the real-time model and offline model configuration pages are different, Chapter 2 mainly describes the way to configure periodic mode, and Chapter 3 will describe how to configure the real-time customer risk rating model.

2. Key benefits

2.1. Advanced Screening with Enhanced Accuracy

With advanced AI capabilities such as multi-dimensional fuzzy matching that perform semantic, phonetic and text matching, we are able to support inner-language matching for all languages and cross-language matching between pairs of over 15 languages, especially languages in the Asia-Pacific region. Coupled with filter function that enables Clients to configure granular filters, we are able to lower false positive rate by approximately 40% and effectively reduce disruptions to good customers.

2.2. Risk-based Approach with Great Flexibility

Appreciate the need to adopt risk-based approach, we enable Clients to effortlessly configure Customer Risk Rating (CRR) rules via a low-code interface that can segment customers info different risk categories with your own rulesets. Results can be passed on to Clients for downstream risk-based activities.

2.3. Powerful Network Analysis Tool with Full Explainability

Our graph-based network analysis tool is able to intelligently uncover anomalies based on reams of data and behavioural/sequential patterns that rudimentary solutions will not be able to handle. Our advanced learning capabilities include few-shot/weak-label/dynamic graph and group detection. All control measures can be explained to regulators with detailed audit trail and also facilitate internal governance.

2.4. Out-Of-Box Rules at Reduced Cost

Thanks to our configurable tools and granular rule settings, Clients are able to perform AML tasks automatically with minimized manual efforts. This is continued with intuitive Case Management portal that allow Clients to review case and generate STR reports effortlessly. For Clients who wish to have a head start, we also offer industry-specific expert rules for Clients to quickly implement into their business. All considered, we are able to reduce up to 80% of manual effort that is typically involved in AML process and significantly saved cost.

2.5. Highly Performant and Scalable SaaS Solution

Tried and proven by over 30 entities and over 300 business scenarios, our cloud-native SaaS solution is highly scalable and performant that can support billions of the transactions per day with close to real-time responses. Clients who wish to start small can opt to adopt pay-as-you-go model that sees significant save on the upfront investment while Clients who foresees potential in expansion of business can flexibility grow their business with us knowing that we are able to deliver the same high standard as they scale up.