Home > Showroom
UNFRAUD ISSUES A CHALLENGE TO COMPUTER CRIME
Avoiding computer fraud is a major priority not only for those who buy online, but also for those who sell. Starting from this basic assumption, three young Italians, Armando Monaco, Vincenzo Paduano and Andrea Puzo, have created Unfraud.
The trio are all in their early thirties and originally from Ariano Irpino, in the province of Avellino, where they first met up as teenagers.
Their new software is capable of recognising an incidence of online fraud and blocking it before it can do any damage. The service analyses a whole series of data and reports the percentage chance that a certain transaction is potentially at risk. It is then up to the seller to decide whether to complete the transaction, to collect more information about the purchaser, or to block the sale.
This start-up has only been operational for a few months, but has already attracted considerable attention due to the innovative nature of its business, and even won some important awards. Technically speaking, the Unfraud service is a hosted solution accessed via APIs and CMS plugins such as Magento, Shopify, Prestahop or OpenCart. The operator of the e-commerce site wanting to check the reliability of an operation enters certain data such as the identity of the customer, the order details, the amount payable and the currency code. Other optional input can be added to this information, which often leads to a more accurate assessment.
The process of analysis relies on a wide set of variables including, for example, geo-location. It may seem obvious, but many fraudulent transactions can be detected by comparing the place where a credit card is used to the place where it was issued, or by seeing if the country chosen for delivery of a product is different from the one in the billing address. These are just some examples of what the Unfraud software analyses in order to detect e-commerce fraud.
After processing the data, Unfraud produces a risk assessment score which ranges from 0.01% to 100%, depending on the probability of fraudulence in the transaction concerned. The score takes into account a multitude of factors, based on all the transactions that the software can detect. Reputational risk data is also used in the process, distilled from the billions of transactions shown in the Unfraud network. The model is designed to recognise errant behaviour and to adapt to new attempts at fraud in real time, using a novel approach based on artificial intelligence.
Among the many advantages this software has to offer is its capacity to be either used alone or in combination with other anti-fraud mechanisms, and the fact that it is virtually invisible to the purchaser. The opportunity to exploit the information collected from across the spectrum of companies using Unfraud, while still respecting the individual privacy of users, enables the system to harness a powerful database, and use it to hone its research and identify potentially fraudulent transactions with more precision.