Aster Data has upgraded its software to speed up data warehousing to improve companies' analytics handling to improve their responsiveness to their customers and to handle online fraud better.
The software which takes advantage of the company's Massively Parallel Data-Application Server to speed up data processing.
Aster Massively Parallel Data-Application Server 4.0 works by embedding application logic within Aster's MPP database, to speed up that processing. In contrast with other data warehousing architectures, data management is kept separate. Applications where this new processing power could be employed include online analysis of shopping patterns, analysis of subscribers within a telco, stockmarket trading analysis and prediction of disease patterns.
According to Aster's Sharmila Mulligan executive vice president of worldwide marketing, the new software was addressing the current trend. "A few years ago everything was about price/performance but in the last 18 months there's been a focus on deeper analytics - in particular, customers want to perform real-time analysis of data, it's not just about storing the data anymore."
Aster's CEO Mayank Bawa said that the new software did more than just speed up processing but provide the possibility of a whole range of new applications. "Impossible queries could now become possible as customers would have the data access they need: before Version 4.0 was introduced these applications would just time out," he said.
As an example, he pointed out that the company had signed up with Full Tilt Power, the online gaming site to improve the company's fraud detection process. Bawa said that use of the software had improved performance from 90 minutes to 90 seconds response time on fraud analysis. "With online fraud, it's speed that's important , It's acting close to the time so that you can collect the evidence. With the Aster software, Full Tilt could detect unusual patterns in the way that hands were played and get that evidence in real-time."
He said that the new version offered other improvements from previous versions of the software. "We've added enterprise-type features such as application process management and dynamic workload management to make it more effective." He added that the company was now talking to SAS about it adding the Aster software to SAS's Enterprise Miner product. "They want to do mining on a large database and we offer them a way to do it."