Overview

A powerful and modern artificial intelligence engine is responsible for transforming data in an optimal and precise predictive model that allows users to explain the causes and effects of a particular fraud. Stands out from other solutions because of its precision and power.

Thanks to its predictive engine, using new data from the transactions of customers to learn and constantly improve its accuracy without human effort. This ongoing process to discover and adapt more quickly to new ways of defrauding that may arise, significantly reducing potential losses.

Capacities

• dVelox EFM is based on predictive models that allow you to "explain" the causes that have a specific type of fraud.
• Tool sensitive to changes in market behavior and adapts to them automatically, detecting new ways of defrauding emerging at a time to anticipate them and make decisions proactively.
• Detect fraud and accurately in nearly real time (on-line).
  Their results support this.
• Detecta el fraude de forma precisa y en casi tiempo real (on-line).
  Sus resultados lo avalan.
  
  • Increase of 35% on average in the detection of fraudulent transactions.
  • 10% of fraudulent transactions were detected previously unknown fraud.

• It feeds with new data coming from the customers' operations. All this in an unattended and automatically.
• The business user creates his own scenarios and simulations, using an intuitive graphical interface lets you work in an autonomous area directly without relying on math or computer science.
• Ease of integration with corporate applications and architectures. Based on standards, to add new functionality as business needs.
• Detection of fraudulent without requiring an extensive historical data. The tool is fed automatically and continuously improving the accuracy learns as time passes.

Value added

• Reduction of losses caused by fraud. Economic losses are reduced by detecting fraud in less time.
• Total coverage of real-time analysis to support the full cycle automated fraud detection. Increases up to 10% of fraud detection so far unknown.
• Reduced operating costs of analysis. The reduction of false positive alerts and allows lower costs of analysis in incidents that not only are not fraudulent but in most cases the rules derived by a downgrade of the same.
• Reducing the costs of creating and maintaining models. Complex and accurate models can be built in hours instead of weeks or months. The immediate result so makes it possible to actuate the mechanisms of prevention and research.
• Ability to react almost in real time to new types of fraud. Compared to other existing technologies based on a reactive strategy, dVelox Fraud Management is able to feed back daily for suspicious behavior at the speed that the market needs.
• Ease of use facilitates greater independence of the personnel against the business rules and analytic tools that require the intuition of a team of experts.