Description of the solution
The Kairos Fraud Detector module is a decision support system based on an analytical engine developed in a BIG DATA architecture, with components for real-time stream analysis (meaning it analyzes data as it flows in real-time), algorithms, and knowledge about criminal activity patterns. Kairos is capable of assessing user risk and proposing necessary actions.
The architecture of the Kairos System allows for the generation of potential breach events in real-time, within milliseconds of receiving an event input. This, in turn, enables the detection of fraud ex-ante, which means before it occurs, allowing for fraud prevention before it happens.
Effectiveness
The system was developed based on the extensive knowledge of a team of programmers and analysts who create IT tools to combat organized crime. Additionally, it leverages machine learning solutions to further enhance the effectiveness of fraud detection.
Executing alerts or control directives to the service provider’s systems about suspected fraud before it occurs.
Experience
A team with experience in implementing BIG DATA based on stream processing with real-time analytics in critical infrastructure systems.
Experience in creating streaming processing systems for data obtained from, among others, smartphone-class mobile devices. Thanks to our experience in co-creating mobile applications transmitting telematics data, we have extensive knowledge of the imperfections of individual classes of equipment available on the market and how to optimize the occurrence of anomalies in the data sent by these devices at the software level.
Analytics
Event analytics in the “input stream”. The system’s response time to detecting a violation/anomaly is milliseconds, enabling fraud detection in real-time and, as a result, blocking fraud before it occurs.
Constant development
Identification of fraud patterns and continuous development of algorithms. Possibility to implement changes without affecting the continuity of processing of the main data pipeline. High availability and flexibility of the system in terms of integration with service providers’ systems.
