Telecoms revenue fraud is a primary driver for increased Apache Hadoop adoption, according to a recent poll of telco and enterprise users by Cloudera and Argyle Data.
Communication service providers lose around U.S. $38 billion to fraud every year and during a recent webinar to introduce Cloudera and Argyle Data’s joint fraud prevention platform, a survey indicated that over 90% of attending organizations already use or intend to use Hadoop for fraud prevention. 34% of attendees said they already have Hadoop in place and may use the platform in their fraud prevention efforts.
“Across all areas of industry, people are trying to figure out the most effective use cases for Hadoop,” said Vijay Raja, solutions marketing manager at Cloudera. “Fraud prevention is a textbook use case for Hadoop-based analytics because the ROI is immediately visible. Real-time machine learning relies on large amounts of data to detect sophisticated revenue threats, making Cloudera the ideal platform on which to run Argyle Data’s threat analytics.”
Cloudera and Argyle Data’s industry-leading native Hadoop application suite uses the latest machine learning technologies against a unique, comprehensive data lake to give communications service providers a 360-degree view of user activities.
“Unsupervised machine learning delivers everything telco fraud analysts need to be efficient at and deliver immediate ROI,” commented Arshak Navruzyan, VP of Product Management at Argyle Data. “The Cloudera-Argyle Data solution interoperates seamlessly with all participants in the Hadoop cluster.
Cloudera and Argyle Data’s next generation native Hadoop fraud analytics platform is specifically tailored for the fraud prevention needs of communications service providers, where sophisticated, high volume attacks drain millions of dollars in revenue within minutes. The platform allows mobile operators the ability to reduce loss by detecting previously undiscoverable revenue threats, delivering up to 350 percent improvement over rules-based offerings.
The platform uses a native Hadoop architecture, combined with real-time data ingestion, analytics, and machine learning. Graphical representations enable fraud analysts to easily see attacks as they happen – a more effective approach than screens full of numbers – which traditional systems would struggle to approximate.