Sign up for The Paypers newsletter Follow The Paypers on LinkedIn Follow The Paypers on Twitter Follow The Paypers on Facebook
The Paypers, paypers, Insight in payments, News, Reports, Events

Cristina Soviany, Features Analytics: "A holistic strategy provides a better understanding of customer`s data"

Friday 8 May 2015 | 11:03 AM CET

The eyeDES modeling technology and execution platform allows us to analyze holistically the customer data, build models per data segment and deliver results in real-time

Could you tell our readers about your background, what has been at the origin of Features Analytics?

After receiving my PhD degree in Applied Sciences from Delft University of Technology, I have led for 7 years the R&D of a Belgium-based start-up, developing a new technology for cancer detection in ultrasound imaging. In 2011, inspired by the principles of modeling used for cancer detection, I have decided to leave the company to initiate the development of an innovative technology to be applied to financial transactions. In December 2011 I have been awarded the prize for leading the most innovative information technology company in Europe.

Features Analytics’ proprietary solution, namely eyeDES, is an innovative machine learning modeling technology and platform for payment fraud detection. By using this technology our team of experts designs highly complex and accurate models that can be immediately deployed in production on the eyeDES execution platform that we also provide.

The omni-channel commerce sector is growing at a fast pace and we are confronted with data coming from multiple sources. How can your proprietary eyeDES platform detect fraud across all levels of the payment industry?

The proprietary eyeDES modeling technology allows us to analyze and understand holistically the customer data and build models per data segment to differentiate more accurately between legitimate and fraud. We deliver a solution that consists of a package of models, each set of models built on a certain data segment. These sets of models deliver results according to a certain set of KPIs. Models with different KPIs can be exchanged in production, in between scoring two transactions in real-time.

It is important to mention that the data segments are identified very often across different business channels and not only within the same channel. We are thus able to deliver the most optimal and stable performance because we can analyze and understand the entire omni-channel data of the customer.

We are very different from other vendors who offer solutions for certain channels or segments of the channel, applying thus a restricted view and solution on the customer data.

We analyze data as a whole and we split it in segments that are modeled separately but in the same time optimized as a whole package. Using our holistic strategy, we provide a better understanding of the business data. This is why we qualify our solution as an intelligent one.

What makes your fraud prevention solution different from similar platforms on the market?

Apart from delivering highly accurate and stable performance of the models we design to score transactions, Features Analytics provides a fully configurable eyeDES platform which allows us to put models immediately in production, shadow mode or A/B testing mode. This capability differentiates us from other vendors who do not have the ability to put the models immediately in production, requiring thus a significant implementation programming effort. eyeDES models produce both scores and reasons which are computed in real-time by the eyeDES platform. The reasons are crucial to enhance the decisions taken in real-time or to streamline the manual review process.

eyeDES platform comes also with a fully configurable dashboard, a real business intelligence tool, to monitor in real-time the performance of the models and any other measures and statistics that are defined by the customer.

What are your plans for the future?

We have just released eyeDES 3.0 which will enable customers to access fraud detection services via cloud. Apart from the cloud version, eyeDES can be embedded in the customer’s current solution or standalone as scalable application server deployment.

We will continue to enhance our innovative machine learning modeling technology and platform, while pursuing important engagements and delivering great service to our customers. We assist them to address the fraud issues, improve their sales, limit their costs and make better business decisions every day.

About Cristina Soviany

Cristina Soviany is the co-founder and CEO of Features Analytics, a new and innovative start-up based in Belgium. Features Analytics provides eyeDES - a fully integrated, configurable and real-time Predictive Modeling and Execution Platform, designed to produce unrivaled modeling accuracy and stability for mission critical environments.

Cristina Soviany has an MSc degree in Computer Science from Polytechnics University of Bucharest, Romania and a PhD in Applied Sciences from Delft University of Technology, in the Netherlands.

About Feaures Analytics

Features Analytics, is an innovative IT start-up based in Belgium, which specializes in detecting payment fraud, with solutions for ecommerce merchants, card issuers, acquirers and payment processors. At Features Analytics we offer modeling tools to build the custom models, the platform to execute the models and score transactions in real-time. The platform contains a dashboard enabling to monitor the model performance against the selected set of KPI’s. Our team of experts design the models, maintain the platform and models, and ensures optimal and stable performance through the time.