NICE Actimize Financial Crime and Compliance

and robustness across the total NICE Actimize portfolio. A key driver of the platform allows for the combination of

advanced rules-based analytics and machine learning AI models wherever they can

be leveraged to improve detection and gain efficiencies in the compliance

process. In that respect, the NICE Actimize platform applies various techniques

where appropriate to highlight specific known scenarios that should be

recognized through a robust set of out-of-the-box rules, using ML models that

learn to classify communications appropriately by intent and eliminate

non-relevant interactions from review, applying NLP to extract meaningful

insights and predictive AI models to highlight changes in behavior of employees.

The platform enables building

sophisticated statistical measures for historical behavior profiling learned in

Batch or R/T, as well as running any kind of ML algorithm or business logic

(some examples include XGBoost, Logistic Regression, Decision Trees, Clustering

as well as Deep Learning NN models). Financial events flow through the system

and then a set of expert as well as data-driven features (risk signals) are

calculated against them, the results and insights derived from the calculations

are captured in a data repository that can be a standard DB or a big data lake

(depending on customer size and preference).

The data is also streamed to the

NICE Actimize analytics lab, X-Sight AI, hosted on the cloud, allowing analysts

and data scientists to analyze the data and create new features and ML models

in a dedicated data science environment named “X-Sight Studio”. X-Sight Studio

is also offered by NICE Actimize to our customers’ data scientists for data

analysis, data visualization and model development and monitoring. Our

customers use X-Sight Studio to develop homegrown models and features using the

calculation produced by the NICE Actimize system, as well as based on

additional data sources of their choice. These models are then integrated to

the NICE Actimize runtime environment where the financial events are being risk

rated for fraudulent activity, money laundering and compliance breaches with

the AI models produced.

Each of the NICE Actimize solutions

is equipped with a set of Out-of-the-Box analytical models and features which

are built using our Analytics development environment and which are being

evaluated throughout the NICE Actimize detection flow. Some of the most common

problems that NICE Actimize solves with AI models are customer segmentation

problems, alert prioritization problems, risk rating transactions for Fraud,

identifying anomalous patterns in trading behavior and identifying risky

behavior from electronic and voice communications.

The NICE Actimize’s philosophy

towards analytics includes a hybrid model of an execution environment (that can

be running on premise or on cloud) flexibly connected to an analytics

development environment that’s running on the cloud.

Data continuously flows from the

execution environment to the cloud environment, which allows NICE Actimize to

leverage the power and speed of the cloud to bring cutting edge innovation to

our customers. This approach enables the NICE Actimize systems to utilize the

most innovative AI techniques and algorithms available at any given moment in

time when models are delivered.

To execute on our strategy of

continuous innovation and delivery, the NICE Actimize analytics organization

consists of a few well-connected groups, each with its own role in creating high

quality analytics.

Each solution has a team includes of

domain experts, developers, analysts, data scientists and ML practitioners that

create and develop features, models, and modeling techniques appropriate for

the specific solution and business problem. Our solution experts leverage the

billions of financial events which flow into our cloud environment every day

from across our customer base to create new solution features, models, and

algorithms. This strategic data asset, which keeps expanding, allows our

development teams to focus on features that bring out significant value in

preventing, detecting and stopping financial crime and test out algorithms and

approaches on real data in a secure and safe fashion.