Analytic Engine

The analytic engine provides a highly flexible and adaptable data model which supports advanced calculations and correlations in real-time. The data model can maintain static and dynamic information on various types of entities including employees, accounts, customers, and others. It can be adapted to specific process and requirements of the organization.

The data captured by the system from various sources is sent to the rule engine in a normalized way, so the platform from which the data has been captured and the way it has been captured become transparent to the analytic engine.

The analytic engine accepts two types of input:

  • User activity events which are identified from the captured user screens which are continuously captured by the system through network sniffing. These events are mapped as “Facts” within the analytic engine.
  • Data from the organization databases and applications log files, for example indication of accounts which belong to celebrities or organization’s executives. This type of indication is typically not displayed on the user screens captured by the system, yet it is important for detecting suspicious behavior.

The system builds a profile for every type of entity, including employees, accounts, customers, ATM devices, debit cards, IP addresses, etc.

The system is highly flexible and enables the customer to build any profile independently, with no need for support from Intellinx. Profiles can be built on any platform and across multiple channels. The system collects data from different channels and consolidates it into one source.

Predefined profile indicators can be adapted to an organization’s specific requirements to signify deviant behavior. Once the system identifies suspicious behavior, it issues alerts in real time, near real time or batch. Real-time alerts can be used for suspending a user or blocking transactions.

One of the unique features of Intellinx is applying new business rules to historic recorded data, which is made possible through the recording and storage of all user activity, regardless of whether it is suspicious or not. This allows internal auditors to check for new potential fraud scenarios in the organization’s pre-recorded data.