RealNetworks is looking to crack down on voice spam and robocalls with the release of an improved KONTXT solution. The new KONTXT for VOICE offering will automatically identify robocalls and other scams that come in through the phone, and drop those calls before they ever get through to an end user.
The solution expands on RealNetworks’ original KONTXT solution, which uses machine learning to distinguish fraudulent texts from legitimate ones. KONTXT for VOICE will do the same for voice channels. It arrives mere months after RealNetworks finalized its sale of Napster, at which point the company indicated that KONTXT would be key a priority moving forward.
KONTXT for VOICE can be integrated into an existing communications platform, and uses natural language processing and audio spectrum analysis to spot robocalls. It also leverages a new voice fingerprinting technique to separate known contacts from potential fraudsters. The solution has already been deployed across two major telecommunications networks, and is expected to help reduce the volume (and the cost) of large-scale voice scam operations.
In that regard, RealNetworks noted that caller fraud has increased during the pandemic. Cybercriminals have tried to take advantage of the disruption with a number of different COVID-19 scams, with many posing as government agents or purveying get-rich-quick schemes. As it stands, Americans currently lose around $10 billion to robocall fraud on an annual basis.
“The KONTXT for VOICE solution goes beyond simple phone number block lists and voice captcha which are easily circumvented by criminals,” said KONTXT CTO Michael Bordash. “Our machine-learning microservices analyze the voice and the intent of the human or robo caller, and make a fast and accurate assessment to help protect global network customers.”
Since selling Napster, RealNetworks has placed a greater emphasis on its biometric portfolio in addition to its machine learning platform. The company recently released a free mask detection app built with its SAFR computer vision technology.
March 24, 2021 – by Eric Weiss