“Trust is the competitive advantage that sets apart winners and losers in the digital economy,” says Jason Tan, CEO of Sift Science. Noting how every real-life transaction depends on trustworthiness, Tan began looking for how to create a similar experience in the digital world. How could online businesses create the best possible experience for their trusted users, without opening themselves up to fraud and abuse? It is with this vision Tan laid the cornerstone of Sift Science, a San Fransisco-based company that leverages machine learning to provide fast and frictionless fraud detection, in 2011. “We democratize the same technology that tech giants use internally to combat fraud and abuse,” Tan adds.

Through APIs, Sift Science’s systems collect a variety of identity and behavioral user data from its customer and run it through a machine learning-powered, automated risk assessment system. The major issues with a legacy fraud prevention approach that center on rigid rules and a team of human reviewers are high costs, inefficiency, and inaccuracy. Sift Science’s Digital Trust Platform leverages machine learning, an advanced technology that prevents payments fraud, account take-overs, content abuse, and more accurately and scalably.

Through its analyst console, Sift Science’s Digital Trust Platform enables organizations to review gray area cases that require additional analysis to determine whether or not they are fraudulent. Users are scored in a 0-100 range, with zero being trustworthy and 100 being extremely risky. This score is accompanied by human-friendly explanations that allow organizations to draw a bigger picture of who is visiting their site or app. “We compute in real time more than 16,000 signals on each piece of data, and we produce the most accurate risk assessment for our customers–akin to a detective solving a crime scene,” Tan explains.