The ability to process megabytes or gigabytes of data in seconds can mean the difference between making more sales or to lose the customer. And in fraud detection, you need engines to crunch data in real-time and detect anomalies and flag them out immediately for subsequent treatment. The whole process has to be a closed loop, not just for handling data, and has the ability to send messages back to control system or other applications for further action. The ability to crunch incoming data, and to connect downstream will have to be built into the entire analytics platform.
Sign up for Computerworld eNewsletters.