Large multinational banks are targets for many kinds of fraud. Chief among these is transaction fraud, in which a fraudster steals someone’s credit card number or credentials and attempts to charge goods and services illegally. For banks to protect their customers and contain losses associated with fraud, it’s necessary to monitor each card swipe, detect unusual or fraudulent activity, and make an immediate decision to allow a purchase to go through – or to block it as fraudulent.
Detecting fraudulent credit card swipes before a charge is allowed to go through requires moving the fraud detection process from a post-facto, batch decision to a real-time, in-transaction decision, a decision that can be made before the card swipe is authorized. Latency is a major challenge, though — consumers will tolerate a few seconds’ delay at most before reaching for a different card, and a lot has to happen in those few seconds — so the fraud detection process itself must be completed in milliseconds. The option of relying on a third-party fraud detection tool to uncover fraudulent transactions can place an organization at risk of monetary loss. Banks need an accurate and trusted solution, one that can ingest card swipe data, analyze it against a sophisticated, continuously evolving fraud model, and make a validity decision in milliseconds.
VoltDB is used to apply bank business rules to a stream of card swipes, hundreds to thousands of times per second. Using stored procedures to manage the analytic logic, paired with state held in VoltDB, banks increase their ability to detect fraudulent cards on the first swipe. As blacklists and rules change, these are uploaded into VoltDB, where the new rules and stored procedures immediately affect card processing decisions.
Banks using VoltDB are able to reduce fraud risk significantly, increasing their ability to detect fraudulent activity before multiple card swipes occur.