Data and machine learning to fight against fraud and non-payment
The solution combines data analysis during the purchase (cart content, delivery method, etc.) with payment data (valid or blocked card, customer's bank, etc.) and the customer's historical and declarative data (address, email, telephone number, identity, etc.). Thanks to web scraping, a technique for extracting content from websites, the data left by customers available to everyone is collected as well as shared banking data (open banking) to study their consistency with the declarative data of the transaction. This open data is used in compliance with the RGPD. A reliability score is thus attributed from more than 400 data. When the score is not really decisive, it is up to the user to decide if he accepts the new customer. He can for example try to contact him by phone to verify his identity. This solution can improve the experience of "good" customers and would divide the risk rate by 4 according to the company. Translated from Levée de fonds de 3 millions d'euros pour la finteech Meelo, spécialisée dans la lutte contre la fraudeOur articles will then appear first in Google Top Stories.
