Passwords are on their way out the door and stronger, more convenient forms of authentication are finding their way into the mobile landscape at an increasing rate as users demand that their smartphones and tablets be treated as their primary network devices. Voice biometrics are particularly well suited for this. Hands-free, unique and convenient, every smartphone is already equipped with the hardware required to measure a voice print.
I recently had the chance to speak with Brett Beranek, Solutions Marketing Manager, Enterprise Division at Nuance Communications (NC) about what makes voice biometrics so ideal in mobility. We begin with some background on the company and its enterprise division, speak about the U.S. Bank pilot of Nuance voice biometric technology for account management, the difference between active and passive authentication, the mobile promise and what we should be looking forward to in the coming year.
Peter B. Counter (MIDW): I was wanting to start off by just getting an overview on the background of Nuance Communications and the general idea of what you do as well, just for the sake of our Mobile ID World readers.
Brett Beranek (NC): Sure. Absolutely. I can start with that. So I’ll give you an overview of Nuance and then I’ll also give you a little bit of an overview about our Enterprise Division which creates the voice biometric technology. Then we can dive into the US Bank application.
MIDW: That sounds great.
NC: At Nuance – to take the 20,000 foot level overview of what we do – we’re all about reinventing the relationship between technology and humans. We’re trying to take that relationship between humans and technology in all the forms that it takes, and make it more seamless, more natural, and more human. Most people have seen the manifestation of that through speech recognition technology. If anyone’s ever used a virtual assistant, or if you’ve ever used a speech enabled vehicle or TV, or what not, there’s a pretty good chance it was powered by Nuance Technology.
So, although the average consumer is often not familiar with Nuance technology – at least not beyond the Dragon software, which a lot of people use at home for dictation – they do use Nuance technology on a daily basis, so any time they speak to their smartphone or any time they use a voice command in their car or what not, they’re using Nuance technology.
The enterprise division is focused on the customer care side of the business, essentially asking: how do enterprises, such as banks and telcos and so forth interact with their customers?
Historically, the Enterprise division tried to make the phone experience with enterprises much more seamless and straightforward by using speech recognition technology, and later natural language technology, so much so that today if you call US Airways, or if you call a number of our other customers you could start the conversation with – or an automated system, and IVR [interactive voice response] system starts a conversation with- “How may I help you?” and then you can say something like “Oh, you know I’d like to change one of my reservations” and then the interaction continues that way. So very similar to that virtual assistant type of experience that we’ve gotten used to on our smartphones.
It’s really taking those core technologies and applying it to the enterprise world, and voice biometrics is one of those key core fundamental technologies that the enterprise division has had in its portfolio for a number of years. And again: with our history, that is technology that’s typically been used in the IVR space – so, in the contact center.
What we’re very excited about – and the U.S. bank announcement is really in line with this – voice biometrics, as well as our other technologies are permeating to the other channels. So, web, but mobile in particular. Voice biometrics is used primarily today within the financial center and within telecoms, those are definitely the two verticals where we are finding the organizations that are using the technology the most. And it’s all about how consumers identify themselves and personalizing that experience.
Instead of passwords and security questions, customers can use their voice to be identified. If we think about some of those experiences that you may have potentially heard about previously (like TD Bank and others) have these systems deployed where you call into the contact center and you say something like “My voice is my password” and you get authenticated through the system and you can perform your transaction once you’re authenticated.
You know, really the U.S. Bank deployment – within their pilot- is taking that same concept but bringing it to the mobile app. Instead of having to type your complex alphanumeric password, you just have to say “my voice is my password,” get authenticated and perform your transactions in a fully secured manner.
MIDW: It’s interesting you touched on how there is starting to be a lot of traction for voice biometrics in the mobile sphere. Would you say that this is primarily due to just the convenience of it all?
NC: I think the primary factor is convenience. If we look at the contact center space, there were a number of factors driving it. One of them was fraud prevention, another was reducing cost in the call center which is really quite significant for long term organizations. Improving the customer experience is definitely a key benefit.
When we look at the mobile app we’re really looking at improving the experience of the user. That is the key driver. And really fulfilling the promise of the mobile app: To be able to perform mobile transactions wherever you are. When you need to type in a password and you’re driving or you’re walking down the street, or you’re in any other highly mobile situation, that can be a challenging experience. And so, what a lot of our customers have told us is that as they study the consumer behavior whenever the user has to enter complex data on the keyboard they need to do so when they are not mobile: when they are sitting at their desk or whatnot. That defeats the whole purpose of the mobile promise of being able to do your functions from wherever you are.
Voice biometrics allow you to bypass that. Voice biometrics, combined with a virtual assistant type capability allow you to perform your functions using speech from end to end.
And that’s really enabling that hands free fully mobile experience.
MIDW. It seems to be something that you are attacking from all angles. To me, you just essentially described why your Dragon Mobile Assistant has pretty unanimous five star reviews on Google Play, and I think that that’s something that speaks a lot to the hands-free experience as well.
NC: Yeah. I’m not sure if you saw the announcement, we brought biometrics to the Dragon Mobile Assistant as well.
MIDW. Yes, we had a news story on that recently. A lot of voice biometric technologies are passphrase oriented, but in the call center space we’re seeing a lot of talk about passive recognition. Can you explain the difference between passive and active biometrics and how Nuance differentiates between the two?
NC: Well those are basically the two voice biometric approaches: one of them is active and one of them is passive. There are different ways of describing it but I think that active and passive makes the most sense. In the active case you need to say something in particular, so it could be a passphrase similar to a password, or it could be something like “Hey, say the following three digits: 574,” or “Say the following random phrase.”
The most common and the most convenient is the passphrase.
The other approach is passive, where you can say whatever you want. So, any type of conversation allows it to listen to your voice and recognize the pattern. In a password replacement type of scenario, so if we’re talking about authenticating a user, the active authentication really is the right way to go. The reason for that is that speed is of the essence. You really want to make that authentication process quicker than if you had to type in a password.
So, that’s one of the key benefits of voice biometrics. And to do that, if we’re going to identify you in one or two seconds of your voice, then by having a pre-defined text when we ask you to speak, we can achieve that with high levels of success.
When it comes to the passive biometrics we need a little bit more audio, in the ideal scenario we would want to have between five to fifteen seconds of your speech. That technology is primarily used in context, like when you are having an actual live conversation with an agent, because that’s where we can get five to ten seconds of your speech. If the contact center agent has confirmation of your identity a few seconds into the call, that’s fine. But, like when you want to get into an app, or when you want to confirm a transaction on the web, or if you want to authenticate into an IVR, you want that authentication process to be as quick as possible. That is where the active technology is best in place.
MIDW: So for the U.S. Bank app, that’s active voice authentication?
NC: Yes, that is entirely the active authentication. So basically, the pilot in their case, they had the user say, at U.S. Bank “My voice is my password,” and that basically authenticated the user into the app.
The U.S. Bank also piloted our Nina personal assistant solution, and so post-authentication then the user can continue with the speech interaction. That can say something like “I’d like to pay my bills that are due in the next week,” or, “How much did I spend last month on travel?”
Those are the types of interactions you can perform with a virtual assistant once the authentication is done.
MIDW: So, essentially, a completely hands-free, secure banking solution.
MIDW: Is there anything we should be watching out for from Nuance in 2014? Anything we can look forward to?
NC: I’m extremely excited about voice biometrics within the mobile space. I think that this is going to be very transformational for voice biometrics. And the reason for that is you know, having voice biometrics in the contact center is great, but most consumers call into their banks and their telco as infrequently as possible.
Let’s say you call once every year, once every two year when you have a special situation, that might warrant a call. Where as your mobile app – especially when we think about banking – that’s something that you use potentially one to several times a month. Whether you’re paying bills, whether you’re checking if a payment has gone through, if you’re looking at past spending behavior – and so all of a sudden with voice biometrics in the mobile app, the average consumer is going to have exposure to voice biometrics on a much larger scale than they have in the past. I think that’s going to be very game changing for voice biometrics as a whole, and obviously for Nuance as well.
We will also be announcing, later on in the year, a number of financial institutions that are deploying voice biometrics in their mobile apps. So that’s definitely a thing to look out for and we’ll definitely get in touch with you once we’re ready to include them.
MIDW: We’re all eagerly looking forward to it. Thanks so much for taking the time to talk to me today.
NC: It was my pleasure.