What if you could identify fraudulent online users before they commit fraud? And what if you could do so in a way that doesn't inconvenience your trustworthy...