ID R&D releases voice biometrics-based user intelligence tech to prevent contact center fraud

ID R&D has released IDFraud Contact Center to prevent fraudsters from using stolen or fake identities to create new accounts and gain access to telco services and equipment, the company announced.

The solution leverages ID R&D’s Text Independent Voice Biometric technology to analyze the voices of new subscribers to identify fraud attempts by comparing voice prints against a database of known fraudster voices.

According to a Europol report, telco fraud is on the rise, generating an annual global loss of some $32 billion.

IDFraud Contact Center is a cost-effective solution that easily integrates with any hardware and software to boost fraud prevention, the company says. It was first rolled out at a telco in January 2020 when integration took one month. The fast fraud identification led to additional deployments in other countries.

The solution’s features include “end-to-end” neural network-based technology, advanced speaker diarization that isolates voices into separate audio files, speed of 1:N identification against a fraudster database, optimal performance with low quality audio, and language and accent independence. Speaker diarization is an advanced use of voice biometrics technology that works in offline mode and provides an accuracy greater than 96 percent.

ID R&D has published a new 12-page white paper that provides a detailed description of the IDFraud Contact Center solution.

“Based on the strength of ID R&D’s voice biometric technology, IDFraud offers a simple approach to a difficult challenge, solving a fraud problem that was not practical to solve in the past,” said John Amein, SVP at ID R&D, in a prepared statement. “Aside from being easy to get started and fast to deliver a strong ROI, Telcos can implement the solution without adding any friction to the new customer onboarding experience.”

Not only does the voice biometric system cross match voices with a watchlist, but it also creates a temporary library of new customers to check if the same voice is used by multiple identities, and estimates gender and age. IDFraud leverages advanced biometric technologies such as ID R&D’s voice biometric product,  IDVoice, which finished with strong results in the 2019 NIST Speaker Recognition Evaluation.

In January, the company released a new version of the IDVoice biometric solution with optimization for edge devices, delivering high accuracy with an AI-based SDK under 1MB in size.

Article Topics

biometrics  |  call centers  |  fraud prevention  |  ID R&D  |  telecom  |  voice biometrics  |  voiceprints