VoicePIN exec discusses new as-a-service voice biometrics authentication solution

January 13, 2016 - 

VoicePIN offers advanced voice biometrics solutions to clients around the world, designed to provide complete security to end users of banks, call centers and interactive voice response help desks.

Using unique voice characteristics, VoicePIN’s voice biometrics technology allows authorized users to access very personal data without having to remember logins or passwords, and lowers the risk of data compromise and identity theft.

Offered as an on-premise and SaaS solution, VoicePIN’s platform is designed for banks, call-centers, insurance industry, healthcare stakeholders and state and self-government institutions, email logins and more.

The company is also a member of a number of industry associations, including the European Association for Biometrics, FIDO Alliance and Biometrics Institute.

BiometricUpdate.com recently had the opportunity to interview VoicePIN marketing manager Karolina Kiepas, who discussed the company’s voice biometrics authentication solutions, the financial institutions it has partnered with, and the upcoming launch of its SaaS platform.

Can you explain how VoicePIN’s authentication solution works?

Karolina Kiepas: VoicePIN compares registered voiceprints. Our technology is based on analysis of the distinctive patterns contained in the human voice. Each unique voiceprint is made of user’s patterns that are stored in the form of statistical models. Verified recordings are stored in a matrix of a set of parameters. Each time a ‘voice’ is being authorized, recording is verified and compared with all previously collected data guidelines. Verification is a process and its aim is to confirm whether the sample of speech belongs to the person who declares access to the data. The result is then compared with the accepted threshold verification. If this number is greater than a con¬figurable threshold, the verification may return positive, otherwise considered a failed process.

What are some advantages that VoicePIN has over other voice biometrics authentication solutions?

We offer voice biometrics technology in two models, depending on what suits best to our client: on-premise or as a service (SaaS). First of all, our system is really fast and simple to install and integrate. Clients or integrators can do it on their own. Moreover, we offer our clients passive enrollment so they can generate fixed passwords from previously recorded data. It’s the most convenient way for the end user, who doesn’t have to be engaged in the process.

Also, VoicePIN operates in three modes of verification: depending on the password, which is a term fixed for all users (text dependent); text-independent and in the Voice Captcha mode where password is generated automatically from a phrase or string of characters that must be spoken. Last but not least, VoicePIN verification of the biometric password is lexical and language-independent so no ASR is needed. Our system can be successfully used in multichannel customer interaction channels.

How is VoicePIN’s biometric system able to block any attempts by imposters to mimic someone’s voice for authentication purposes?

VoicePIN technology is resistant to voice mimicking due to the nature of the intrinsic algorithms used for biometric analysis. Biometric traits which are utilized by VoicePIN solutions are not what people consider to be the perceived “sound of voice”. Those traits used for voiceprint registration consist of much more information than a human can perceive and intentionally mimick. It was proved by tests performed at AGH University of Science using the VoicePIN technology, in which VoicePIN was 100% robust to voice mimicking attempts, while human listeners cohort was confused in up to 70% of cases. The tests were performed by professional voice talents and voice actors with an extensive knowled ge of the speech production process and with the support of the technical speech monitoring tools. Moreover, VoicePIN system is designed to protect from hacker attacks. It also prevents deceptions by blocking access to accounts for cyber thieves. VoicePIN is equipped with Active Playback Detection so it can parameterize recordings and a statement of its quality. Our system offers detection and monitoring of intrusion attempts based on recordings database and adapt itself to the user variable intonation.

Who are some financial institutions VoicePIN has partnered with?

Last year we covered a huge implementation in one of the European Ministries of Finance, consisting of 10,000,000 voiceprints. Our voice biometrics system also works in Smart Bank’s mobile application for both business and individual customers. They no longer need to enter passwords or PIN codes to make a transfer or check their balance. VoicePIN technology was successfully deployed in a few other European banks and insurance companies. We’re now in the middle of several POCs (proofs of concept) but we’re unable to reveal more information at this stage.

In terms of new product developments and partnerships, what does VoicePIN have planned for in 2016?

Our roadmap for 2016 is really impressive – we’re planning to enrich our solution with new functionalities and, primarily, launch our SaaS platform which is almost finished. After we got to Top 10 Startup Alley companies at TechCrunch Disrupt in San Francisco (held September 21-23, 2015), our office in the USA is growing so this market will be really important in the next several months. Also, we will show our products at a few important events like Finovate Europe, SpeechTEK and Mobile Voice conference so we’re hoping the next year will be full of successes.

Leave a Comment

comments

About Justin Lee

Justin Lee has been a contributor with Biometric Update since 2014. Previously, he was a staff writer for web hosting magazine and website, theWHIR. For more than a decade, Justin has written for various publications on issues relating to technology, arts and culture, and entertainment. Follow him on Twitter @BiometricJustin.