Nuance Communications’ introduction of the virtual assistant named Nina marks a significant set of advancements for mobile customer care. Nina’s software and approach starts with a novel combination of voice biometrics, speech recognition, text-to-speech rendering and natural language understanding, all hosted in Nuance’s “cloud.” The offer comes packaged either as a pre-configured “persona” (named Nina) that can serve as a speech enabled guide for an existing mobile customer care application, or as an SDK (Software Development Kit) that provides enterprise IT personnel or 3rd party application developers with access to an API (application programming interface) to speech processing and computing resources, reference models and software development tools that enable them to customize natural language access to their customer care services through a virtual personal assistant.

Comparisons to Apple’s Siri are inevitable, so here goes. As an homage to Star Trek, the TV series that made talking to computers seem so natural, Siri and Nina have very different “Prime Directives.” Siri responds to utterances about a broad set of personal tasks for iPhones (and, soon, other iOS based devices). Its strengths are initiating phone calls, taking dictation, finding music to listen to, and answering simple questions. Lately, it often responds by asking “Would you like me to search the Web for you?”. That signals that its designers have realized that you can’t be expected to anticipate, understand and know the answers to every question the general public may generate. As individuals have become more familiar with what Siri can and cannot do, they are subtly changing what they ask of it. In the mean time, Apple will use the collective utterances of Siri users to define the services it offers and refine the ways it fulfills requests, most often mapping to the resident address book, local maps, and Web-based search.

By contrast, Nina is part of an enterprise’s customer care offering. Because it already knows what company an individual is calling, it will have a pretty good idea of the purpose of the call (based on call history). It will also benefit from analysis of other data and metadata to understand the limited number of options an individual is likely to pursue. But Nina, by design, goes one step step further. By incorporating voice biometric-based speaker authentication, it can have high confidence in the identity of the caller. Once it knows who you are and the company you are calling, it can arrive at why you are calling and how to resolve your problems in relatively short order. That spells higher probability for a better customer experience.

Because Nuance offers Nina as a service or as an SDK, applications developers can tailor the Nina experience to fit specific business requirements. Insurance and financial services specialist, USAA, will be the first to put Nina into service as an assistant to its increasingly mobile roster of military veterans and active service personnel. The company has always put a premium of innovative customer care and was one of the first businesses to integrate natural language understanding as part of its self-service call flow. It will launch a pilot in August, but intends to make Nina an integral part of its popular mobile customer service application, flavors of which support banking, investing and both auto and property insurance). The USAA Mobile app is available for iPhone®, iPad 2®, Android™ and Windows® Phone 7 devices.

Nuance is taking an “open” approach to offering an SDK, in that it is available for download to developers with an interest in building a speech enabled virtual assistant and seeking to leverage Nuance’s know-how and pre-baked apps. That differentiates it from Apple’s locked-down approach to controlling the customer experience. Few developers expect a Siri API to be issued in the near future. Still, Opus Research believes there to be more than two dozen firms aiming to leverage expertise in customer care analytics, artificial intelligence or automated chat with an eye on revolutionizing self-service with virtual personal assistance. This is where VirtuOZ, Anboto, NextIT, Expert Systems, Artificial Solutions and VoiceBox might prove to be very interesting competitors (or partners) to Nuance and Nina.

When all is said about (and to) Nina, we believe that the integration of voice biometrics-based speaker authentication will emerge as its key differentiator. As my colleague Greg Sterling discusses in his post on Internet2Go, password- or knowledge protection of all things mobile is turning into a nightmare. Voice-based authentication seems like the obvious solution. Baking speaker authentication into a virtual assistant is a very good idea.

Categories: Articles

Tags: Natural Language, customer experience, Virtual Assistants, Nuance, mobile speech