Pindrop announces Deep Voice machine learning technology

Pindrop today announced their proprietary Deep Voice biometric engine, which utilizes the latest deep neural network technology for speaker recognition, to passively identify legitimate and fraudulent callers, solely by voice.

Pindrop’s Deep Voice engine has expanded the company’s total addressable market from $2.5B to $7.8B with the introduction of a new authentication product, Pindrop® Passport.

Pindrop Passport is a multi-factor authentication solution that passively authenticates callers based on their voice, device and behaviour as they naturally engage with a call centre. It is powered by Deep Voice biometric engine that is able to recognize a caller by their voice even with short command-like utterances, regardless of if they are interacting in the IVR or with an agent.

Vijay Balasubramaniyan

Vijay Balasubramaniyan

“Our newly patented Deep Voice engine is creating new opportunities for Pindrop as we see voice eating all other interfaces,” said Vijay Balasubramaniyan, CEO and co-founder of Pindrop.

“Establishing trust, security and identity is key to unlocking brand loyalty with consumers. With today’s product expansion, we are delivering an end-to-end voice identity platform for enterprises looking to reimagine the customer experience as voice becomes the dominant interface of choice.”

Pindrop’s patented Deep Voice biometrics was developed to address a number of shortcomings with the existing systems in the market such as poor enrollment rates and accuracy that are further exacerbated by short utterances and poor call quality. The Deep Voice engine also brings enterprise-grade voice security by detecting new voice attacks such as voice synthesis and distortion and will usher in a new era of enterprise virtual assistants, chatbots and frictionless customer experiences.

Passport is comprised of three core machine learning technologies: Deep Voice biometrics engine, proprietary Phoneprinting technology and Toneprinting behavioural analytics technology. Passport runs passively in the background of every call and identifies a caller based on who they are based on their voice, what they have based on their device and what they do based on their behaviour.

This replaces weaker forms of authentication that currently exist, including asking customers a set of knowledge-based questions or laborious enrollment processes. Enterprises can now reduce call handle times by up to 55 seconds and call centre operations cost by up to $1.00 per call, as well as reduce fraud losses and customer churn.