CST: VoiceKey.Platform

The main articles are:

• Biometric identification technologies
• Speech technology: On the path from recognition to understanding
• Fraud Detection System

VoiceKey.Platform is a multi-modal biometric authentications user platform in remote service channels, adapted to work in,, IVR contact centers web and mobile applications.

In contact centers, including IVR, the biometric platform CST allows you to implement identity verification by voice, in web- and mobile applications - by voice and face. A unique feature of VoiceKey is the patented "live user" detector, which excludes the possibility of using surrogates - voice and photo recordings - for verification.

The platform also solves the tasks of storing data, delimiting access, scaling and ensuring fault tolerance. The platform provides a programming interface for user applications. Thus, biometric authentication from the CST is easily integrated into the software of banks, medical institutions, contact centers, any mobile and web applications. The customer can determine the business logic and use case on his own, or he can use finished CST products, for example: VoiceKey.AGENT to verify users of contact centers, VoiceKey.FRAUD to combat fraud in the telephone channel or OnePass cloud service.

2022: Development of user interface, voice and facial anti-spoofing

Speech Technology Center Group introduced an updated version of the multimodal platform. biometric authentications This was announced on June 17, 2022 by the general director of the CST group of companies. Dmitry Dyrmovsky The project was implemented with support. Russian Information Technology Development Fund

VoiceKey.PLATFORM is a platform that combines products and services to identify customers and prevent fraud in remote service based on voice and facial biometrics. The main features of biometrics are improving security and improving the quality of customer service. The decision has already proved its effectiveness in banks and financial institutions in Russia and the CIS, the platform is in demand in the foreign market. An updated version of the biometric platform may be in demand in banks, retail, MPSC, telecom, development and other industries.

The principle of the platform: when calling the contact center, at the moment of starting a conversation with the operator, a user check is launched - a voice model is created in real time, compared with the standard and monitors the appearance of the voice of strangers, including those previously compromised. The whole process takes a few seconds. The result of confirmation of the identity by voice appears on the screen of the contact center operator's monitor. With the introduction of such a solution, you do not need to remember the card number, passport, code word or other identification information: to obtain personal information on your accounts or operations, only a voice is enough. In this case, the artificial voice system instantly recognizes and informs that the verification has not been passed. The solution does not depend on the language in which the client speaks. All this makes biometrics not only more convenient than traditional methods of confirming your identity, but also many times more reliable.

Thanks to the support of RFRIT, it was possible to implement a number of global platform updates in a short time: to improve the user interface, develop voice and facial anti-spoofing technologies (protecting biometrics from hacking).

Voice biometrics examines special voice characteristics and "recognizes" a person by voice, which is more reliable than, for example, a request for personal data - please name the last numbers of the passport or the last complete banking transaction. To pass voice biometrics, scammers can play a recording of a person's voice or go a technologically more difficult way - fake the desired voice using Deepfake technology. To detect such actions, Antispoofing technology is used - anti-hacking protection.

The R&D team of the CST group, within the framework of the project, developed methods (algorithms) for detecting various types of attacks on voice biometrics.

The subject of the study was three scenarios:

• Detection of attacks based on replay. In an attack of this type, the fraudster receives a recording of the user's voice in some way, and then plays this recording when trying to hack.

• Detection of attacks based on synthesis or conversion of voice. In such an attack, fraudsters can use arbitrary samples of the user's speech (for example, from audio files available on the Internet) and use a cloning or voice conversion system to obtain synthesized speech to hack a biometric system.

• Detection of Deepfake attacks, where it was proposed to distinguish real voices from artificially created voices in difficult acoustic conditions.

The CST group was able to achieve high results in protecting biometrics by using neural network architectures, approaches to their training, as well as non-standard ways to augment training data (making changes to data in order to increase their variability and obtain a more stable algorithm).

2019: Compatibility with Alt 8 SP OS

On July 18, 2019, BASEALT"" announced that together with "CST innovation "(MDG GC) they successfully completed testing for compatibility of the latest versions of their products - biometric authentications the multimodal VoiceKey.PLATFORM software complex produced by" CST innovation " operating system Alt 8 SP and the development of" BASEALT. " The correctness of the collaboration of software products, confirmed by the certificate, allows you to recommend them for building reliable channels of remote customer service. The developers are confident that the software package will find wide application public sector in business.

The compatibility of the solutions "BASEALT" and "CST innovation" will allow customers who have begun the process of import substitution, for example, the public sector, to use biometric technologies. We see wide prospects for building on the domestic technological platform a variety of public services with a high level of reliability, accessibility and security for personal data of citizens,

The presence of a compatibility certificate is also interesting for our "classic" customers from the financial sector and. Taking telecom into account the trend towards import substitution and in commercial structures, we will be able to recommend Alt 8 JV to customers interested in installing VoiceKey. This will allow us to implement biometric projects both for and contact centers for collecting samples Unified Biometric System RUSSIAN FEDERATION based on our compatible solutions,

2017: Support for monochannel operation

The updated version of VoiceKey.PLATFORM implements support for the monochannel mode of operation (diarization). The function allows you to separate operator and client speech in real time to simplify the further work of security specialists. The cross-channel mode of text-independent voice identification has also been improved, which will allow you to register and verify on various devices with higher accuracy. And to facilitate the work of developers, a public software interface based on the REST-API protocol has been implemented.

2016: VoiceKey.PLATFORM biometric platform from the CST is included in the Register of domestic software

VoiceKey.PLATFORM, a multimodal software package biometric authentications developed by the "MDG Innovation" (resident, Skolkovo CST group Ministry of Communications and Mass Media of the Russian Federation), was introduced by decision in November in. Unified Register of Russian Programs The register was created in accordance with Article 12.1 of the Federal Law "On Information, Information Technologies and Information Protection" and contains information about all software that is officially recognized as originating from. Russia

2015

CST introduced the VoiceKey 2.0 biometric platform

VoiceKey 2.0 is used to authenticate clients in remote service. Confirmation of identity using a voice and a person allows you to abandon PIN codes and passwords.

VoiceKey 2.0 is secure access to a mobile application or personal account on the site, protection of personal and corporate information on a PC or in a mobile application, convenient verification in a contact center, including in. IVR VoiceKey reduces the risk of fraud and reduces customer service time by applying multifactor authentication where it was previously impossible - in a telephone channel.

The biometric platform from the CST is distinguished by support for voice and facial biometrics with the ability to connect other modalities (finger, palm, iris, etc.), scalability, linguistic independence, flexible integration with mobile and web applications, most popular PBXs, IVR and CRM, the ability to configure access rights.

Authentication is carried out by passphrase (static or dynamic) and by free speech, online and offline modes. Moreover, the accuracy of verification by passphrase is increased by 2 times. VoiceKey 2.0 has a built-in audio pickup server, added virtualization support and a new security engine.

VoiceKey.Agent 2.0 - voice authentication solution in CC

The Speech Technology Center presented in March 2015 a generation solution for background confirmation of the client's identity during a conversation with the contact center operator. VoiceKey.AGENT 2.0 reduces verification time by 50%, works for incoming and outgoing calls and checks the identity of the client simply by voice, regardless of the spoken phrase.

To register a client with the system, you only need one call to the contact center. VoiceKey.AGENT 2.0 remembers its voice parameters and creates a benchmark for comparison. On subsequent calls of this client to the contact center, the verification process starts automatically: VoiceKey.AGENT 2.0 compares the subscriber's voice with the standard stored in the system memory and displays the verification result on the operator's screen.

The scalability of the system (from 10 to 10,000 operators, from 100 to 10 million reference standards for comparison) allows you to use VoiceKey.AGENT 2.0 in both small and large contact centers serving banks and telecommunications companies. Integrating VoiceKeyAGENT 2.0 with your existing contact center infrastructure is easy. The solution has become cross-platform (Windows, Linux, virtualization support), text and language independent (comparison of voice parameters does not depend on the client language and spoken phrase), compatible with telephony platforms of leading manufacturers (Avaya, Cisco, Alcatel, Genesys).

VoiceKeyAGENT 2.0 is embedded in the interface of an existing CRM one and allows you to exchange data with it. The operator does not need to work in an additional application, so the implementation biometric authentications does not require significant changes. business processes

2014

Confirm Customer Identity by Using Voice

As of July 2014, VoiceKey offers a solution to confirm the identity of the client by using the user's voice as a single password in any system. At the same time, voice can also be used to effectively detect fraudsters among callers at a contact center.

The VoiceKey platform provides the ability to conduct a user verification process in the automatic or background mode, which significantly relieves the load on contact center operators, and reduces organization costs.

VoiceKey allows you to make the process of remotely contacting a client to a company easier, faster and more convenient. Now customers do not need to keep in mind a large amount of various auxiliary information (passwords, code words, etc.) The VoiceKey platform uses unique voice characteristics to confirm identity regardless of the selected remote access method (by phone, through web resources, specialized applications on smartphones, etc.).

VoiceKey works in a mobile bank

Integration with Avaya

February 19, 2014 it became known about the compatibility of the voice biometric platform VoiceKey with key solutions of the Avaya corporate communications system for contact centers.

The Voice Technology Center development company said that all VoiceKey products have been successfully tested at Avaya Test Lab and are certified to share with Avaya Aura Experience Portal# 6.0 and Avaya Aura Communication Manager# 6.3.

"The CST is a leading developer of innovative solutions in the field of high-quality recording, processing and analysis of audio-video information, speech synthesis and recognition and bimodal biometrics, prides itself on the fact that, thanks to the integration of VoiceKey with Avaya, confidential information of millions of users using the services of contact centers working on Avaya infrastructure will be under reliable protection," said Ivan Veresov, executive director of the CST.

VoiceKey will provide secure access to Fujitsu and IWS cloud platforms

Fujitsu Corporation signed a technology partnership agreement in the summer with one of the leading developers of ImageWare Systems identity verification systems. Now the IWS biometric user identification solution, based on CST 's VoiceKey technology, will provide secure data access in Fujitsu's NuVola Private Cloud platform.

The Speech Technology Center and ImageWare Systems have been collaborating since spring 2014. The first joint project of the IWS and CST was the IWS pillphone mobile drug management application with built-in biometric user verification. To access your health information, you need to go through VoiceKey voice and face identification.

By becoming an official partner of ImageWare Systems, the CST has also integrated its technology into IWS cloud solutions for multimodal biometric verification of GoCloudID and GoMobile Interactive.

The high quality of the CST biometric algorithms is also confirmed by the victory in the i-vector challenge 2014, organized by the American National Institute of Standards and Technology (NIST).