Sift

Sift is a fraud prevention platform that uses machine learning trained on a large cross-customer data network to score the risk of users and actions in real time. Companies send Sift events from their site or app, such as account creation, login, order, or content post, and Sift returns a score indicating the likelihood of payment fraud, account takeover, or abuse, which the business uses to block, allow, or review the action, backed by a console where analysts can investigate a user's full history and tune decisioning rules. Because Sift sees signals from thousands of sites, it can recognize a device, card, or pattern that has committed fraud elsewhere even if it is new to a given customer. Its products span payment fraud, account defense, dispute and chargeback management, and content and marketplace abuse, and it has introduced frameworks that combine identity signals across dimensions to improve accuracy. Sift is used by hundreds of global brands across e-commerce, fintech, travel, and digital services, and it publishes a regular index on fraud trends. It is sold as a paid platform, typically priced by the volume of transactions or events scored, with packages by use case.