Ravelin Risk Prediction
Any business that sells goods or services online is vulnerable to attack by fraudsters. It is the most common crime in across the globe in every advanced market. The traditional approach to tackling payment fraud, is to use rules or logic statements to query transactions and to direct suspicious transactions through to human review. This is slow, expensive, and no longer suitable for this new evolving world of real-time purchasing. Remarkably it is still the strategy of most of the leading payment gateways and acquirers, a great many of them Dutch-based.
Ravelin identified a far superior approach, which is to extract patterns of fraudulent behaviour using machine learning. By integrating with a business’ website or mobile app to get a real-time feed of customer data, a machine learning model can provide the client with probabilistic scores of the likelihood of the customer being fraudulent. Machine learning allows us to do this in less than 400ms guaranteed. It also allows us to scan thousands of transactions every second, scaling to millions of fraud scores per month. This is vital to our market-leading clients that include Deliveroo, Just Eat, MyTaxi, Kinguin, eShopWorld, and many more.