STOCKHOLM -- TeliaSonera International Carrier (TSIC) today announces the implementation of an automatic proactive fraud detection service for the benefit of its international voice customers. This service enhancement follows development and tests with IPsoft, a leading provider of fraud management and big data application services.
As international voice services move away from a physical circuit switched environment, detecting new methods of fraud such as deliberately extending call duration, or sending calls to automated message platforms, becomes an increasingly important issue. TSIC is taking steps to protect its customers and their end users using IPsoft functionality which performs statistical analytics to identify the source of the fraud, enabling TSIC to take action and maintain the integrity of its service.
“TSIC is committed to providing the best customer experience for our international voice partners both today and in the years ahead. Promptly identifying and taking action on fraud is key to security and maintaining the integrity of our services.” said Simon Dodsworth, Vice President, Head of Voice and Mobile, TeliaSonera International Carrier, “The solution implemented with IPsoft is a major step in our fraud detection capabilities and we believe it moves us to the forefront of carriers offering fraud avoidance solutions while delivering an enhanced level of service for our customers.”
“We originally developed our autonomic system to learn the patterns of falsely answered calls for major enterprises and have been very pleased with the interest shown by TSIC in tackling the problem at source.” said Ravi Dube, CTO at IPsoft, “We had a very successful trial which demonstrated that the non-invasive technology worked perfectly, finding fraud without false positive alarms and its results underpinned a solid business case.”
The solution uses a self-learning set of statistical and big data algorithms developed by IPsoft. False answers are detected and patterns indicating fraud are detected in real-time. This is done by analyzing parameters in call detail records (CDRs) to monitor deviation from statistically ‘normal’ behavior by route. The solution returns 98% accuracy levels and scales to 500 million calls per day.