Interview With Alexey Khitrov, President, Speechpro

Peter O’Neill, President, findBIOMETRICS (fB): Can you please give us a little bit of background on the company?

SpeechProNLAlexey Khitrov, President, Speechpro (SP): Speechpro is the North American subsidiary of the Speech Technology Center. The company was founded in 1990 and has been around for over 20 years. Two key factors separate us from other companies. One is our emphasis on R&D. We do have probably one of the biggest R&D capabilities in voice biometrics with about 150 people in R&D alone (total staff is 400) including 30 PhD’s on board. The other thing is the breadth of the offering and the breadth of the technological expertise that we can bring to the customers. We have products in everything audio from multichannel recordings, speech analytics, signal processing, both text-dependent, text -independent voice verification (1:1) and voice identification (1:N). So the focus on R&D and the breadth of our offering are the two things that separate us from the rest. We have a long history in law enforcement segment of voice biometric market (including some of the world’s largest deployments) and recently introduced our solutions to the Enterprise customers.

fB: Can you tell us a little bit about one of your latest product releases, the VoiceKey.OnePass?

SP: Absolutely. So VoiceKey.OnePass is an innovative idea that brings together different biometric modalities into a single verification process. The idea here is to make verification in the mobile environment or online as easy as possible while increasing the security. So what we are doing is combining the modalities in the collection process. We are asking a user to say a password, while the password is being said a picture is being taken by the smartphone/laptop/camera phone. That is why the technology is so popular right now because if you look at the market space there are over a billion smartphones in the hands of the customers right now and we are expecting over 3 billion in the hands of the customer by 2018. That is the true size of the opportunity. OnePass utilizes capabilities that are already embedded in the hardware that is in the hands of the customers unlike in the case of the fingerprint which requires additional scanners,. It is not only a very easy verification process  but the accuracy is much higher than what is typical with a single biometrics solution. OnePass also provides and extra layer of security with the liveness detection capability which brings together the facial recognition and the voice biometrics technology as we are tracking the lip movement of the speaker while the password is being said. The concept is simple, in order to say the password we must move our lips, so we are tracking these movements to make sure it is a real person in front of the device and not a photo. When we look at the history of applications that allow access the secure content in the mobile environment or on the web we see that it has been like a tug of war between the customer experience and the security. In order to make things more secure the customer experience would suffer or to make the customer experience more fluid the security would suffer. Here we are ending this tug of war.

fB: I think you have also answered my next question, which was voice biometrics is really gaining ground especially in the mobile environment, what are the key drivers? You have mentioned a couple of the drivers, what else do you think is happening in the marketplace right now?

SP: Well I think the role of the mobile is increasing in human life so we are doing more and more with our mobile devices and this is a great opportunity for biometrics in general and voice biometrics in particular as a microphone is an integral part of any phone. Voice biometrics can do is provide secure access to the protected resources on the device, on the server or through call centers. Of course we see a lot of potential in multimodal biometrics in combining biometric modalities and getting the most of security without depletion of the customer experience. The fact that the computation capability of the smartphones is increasing so rapidly allows us to add more and more biometric capabilities right on the smartphone. This trend creates fantastic opportunities for the marketplace and we already see a lot of growth in this space.

fB: You’ve mentioned law enforcement and mobility, what other vertical markets are you focusing on?

SP: We are looking at the segments like financial… that is a big driver, healthcare, government services; these are probably the key drivers at this point. But essentially in any type of protective environment, be it a protected corporate network or be it a paid content or a phone as a wallet, they all require some type of authentication and if biometrics can provide easy to use and secure verification we are in. The opportunity is tremendous because fundamentally biometrics as an access control is ideal. The technology is “disruptive” because when you compare it to the existing password based verification, biometrics are just more secure. It’s not “what you know”, but “who you are”. Passwords can be stolen, they can be intercepted and someone can look over your shoulder. Most of us have multiple accounts on social networks, with banks, companies and so forth and by human nature we tend to repeat our passwords. So if somebody knows your Facebook or Skype password, chances are you are using the same password in some of your other accounts as well. Now, your security is compromised. When you are using biometrics that is a different story, your voice or your face cannot be stolen.

fB: Can you describe other trends in voice biometrics for us please?

SP: We have done a number of very exciting deployments both in the mobile space and in the more traditional voice biometric call center space and I want to mention a new product that we are introducing to the market right now. It is called the VoiceKey.Agent. It utilizes a technology of text-independent voice verification. In this opt-in system the customer has agreed to be verified by voice biometrics and the next time a customer calls the call center, the agent would see a notification on the screen and continuously throughout the conversation would be getting a confidence score that the agent is talking to the right customer. Without any use of passwords, without any effort from the user whatsoever the identity of the user will be continuously confirmed for the agent so the change of information or the transaction in this environment can be absolutely secure. And to make it even more secure we have a feature that detects a speaker change, so if the speaker at the other end of the line changes the agent would be notified.

fB: Has this been a good year for Speechpro?

SP: Yes it has. We had some very exciting new products. We talked about OnePass, we talked about Agent and some of the other products as well but we are even more excited about 2014. The amount of projects and the amount of deals that we are carrying into 2014 is unprecedented for us.

fB: Thanks so much for taking the time to speak to us today.

SP: It has been my pleasure Peter.