White-label acquiring platform Dimebox has introduced a new fraud predictor service based on machine learning.
The service is designed to train itself by assessing batches of transactions, both fraudulent and legitimate, to develop the pattern of frauds targeting individual merchants.
Go deeper with GlobalData
By identifying evolving fraud patterns, the system is capable of blocking fraudulent payments allowing only legitimate transactions.
It utilises self-learning algorithms to determine the fraud score of every transaction. Based on the fraud score, the user can block the transaction.
Furthermore, the Dimebox platform allows setting transaction rule sets to enable blocking fraudulent transactions before they are completed. It will automatically block a transaction, if the determined likelihood of fraud is above a defined limit.
Accordingly, the filtering process of the transactions can be adjusted precisely based on the user’s requirements thereby enabling the perfect balance to allow authentic transactions and blocking the fraudulent ones.
The Dimebox full-stack white-label acquiring platform provides end-to-end processing where all fraud data are directly procured by the gateway through chargebacks and fraud reports such as Visa’s ‘TC40’ report and Mastercard’s ‘SAFE’ report.
