Features Analytics Launches eyeDES 4.0 Machine Learning Risk Management Platform for Better Accuracy & More Stability

Belgian company Features Analytics will be present at the Money 20/20 startup exhibit hall this October in Las Vegas. Co-founder and CEO Cristina Soviany will present the new Features Analytics’ eyeDES 4.0 machine learning solution. The eyeDES platform for scoring risk and fraud is equipped with powerful automatic model retrain functionality, adding more resilience and user-friendliness and is aimed at large financial institutions and large brand merchants.

Features Analytics' patent-pending eyeDES modeling technology and platform utilize principles stemming from ultrasound tissue characterization technology used for detecting cancer. At first, the migration from identifying complex life-threatening cancer cells to detecting risks or fraud for financial organizations may seem like an unusual leap, said Soviany. With billions of dollars being lost to fraud annually and declining consumer trust, it is obvious that companies need to adopt new state-of-the-art machine learning technologies. And even more so as fraud evolves continuously therefore increasing the risk, putting businesses and brands in danger.

There are a lot of similarities between detecting cancer and fraud. Both have complex and large data sets that require deep analytics, sophisticated machine learning algorithms to detect abnormal cells or suspicious behavior. Both require the highest precision in diagnosis, importantly to reduce false positives. You must gain the trust and confidence of the consumer/patient with your course of treatment and get the right answers in real time. Precision and highest confidence are at the core of our company's vision, said Cristina Soviany.

Features Analytics offers a truly complete, streamlined and configurable solution that achieves:

  • Accurate scoring results with detection rates up 30%
  • Confidence necessary for real-time decision-making
  • Reduced number of false positives, by at least 20%
  • Increased stability of models and predictions

Features Analytics builds and supports multiple custom predictive models, each tailored specifically to the risk tolerances and business objectives of the organization. They are responding to the specific requirements of each business, scoring above 3000 transactions per second in real time.

With the eyeDES streamlined machine learning platform, the sets of predictive models can move instantly from design to execution in production, reducing the need for costly and time-consuming programming drastically.

The platform provides a user-friendly dashboard, a real management tool leveraging artificial intelligence capabilities, enabling control across multiple channels, models switch on the run, parallel execution in production, shadow, A/B testing mode and other operations.

See the link to the schedule of Features Analytics’ presentation below:

About Features Analytics:

Features Analytics is a Belgian company providing eyeDES, an advanced machine learning modeling technology and streamlined execution platform for event risk scoring and preventing payment fraud, offering solutions for e-commerce merchants, card issuers, acquirers, banks and payment processors.

You can contact Features Analytics by visiting them at Money 2020 Startup Row in the exhibit hall or by visiting the company’s website at For direct contact, email Richard Mener at

For more information, please visit us online at

About Money20/20:

Money20/20 is the largest global event enabling payments and financial services innovation for connected commerce at the intersection of mobile, retail, marketing services, data and technology. With 10,000+ attendees, including more than 1,000 CEOs, from over 3,000 companies and 75 countries expected at its 2015 US event, Money20/20 is critical to realizing the vision of disruptive ways in which consumers and businesses manage, spend and borrow money. The next Money20/20 will be held in Las Vegas, October 25-28, 2015, followed by Money20/20 Europe in Spring 2016.