There’s a moment in the movie “Her” when it becomes clear that the protagonist’s virtual assistant (the eponymous “her,” herself) is clearly conversing with other individuals and cyber-beings. It is much more than a plot twist in the movie. It is a consciousness-raising moment for all the movie viewers. It showcases a schism, of sorts, between the personal virtual assistant that is embodied (or embedded) in a mobile device and a broader community of “intelligent assistants” that have garnered and will continue to gain deep understanding and knowledge of discrete subjects, vertical industries and areas of concentration.

I was reminded of this in the course of a briefing with Dror Oren, VP of Product at  Kasisto, the SRI spin-off that is making great strides into the banking and financial services vertical. Less than a year since it raised seed funding and left the SRI fold, the company has refined its product and service offerings and reports significant progress with banks around the world, reporting several paid proof of concepts underway and formidable steps toward large-scale product launches before year’s end.

During eight months of marketing and implementing solutions to banks, the company has refined its product offering, positioning and marketing message. It asserts that Siri, Cortana and GoogleNow will have broad application as intelligent assistants to mobile device owners, but only limited depth and breadth of understanding of how to recognize and fulfill their objectives and intents when dealing with specific companies or vendors. Companies, on the other hand, benefit when they can enhance their mobile applications with the ability both to understand and to respond quickly to their customers’ requests by understanding natural language.

Kasisto makes the distinction between “Personal Assistants,” like Siri or Cortana and “Personal Specialists,” which is the role that Kasisto plays for banks. This is a profound observation; and one that will help us define a big segment of the Intelligent Assistance ecosystem. Artificial Solutions was one of the first Intelligent Assistance “pure plays” to recognize the need for automated resources to deliver domain specific knowledge in response to user queries. It introduced the Teneo Network of Knowledge in February 2014. Teneo is ArtSol’s core Natural Language Interactions platform and it was demonstrated running on Indigo, the company’s mobile intelligent assistant.

By running the Teneo Network of Knowledge in association with Indigo, ArtSol is able to show how a “personal” assistant can identify and defer to specialists that are more knowledgeable about a particular topic. In the real world, those subject matter experts are likely to be affiliated with a particular company, brand or automated intermediary (like Amazon, Expedia, Orbits, Hotels.com).  Siri, for instance, was originally conceived to generate most of its revenue by completing purchases of movie tickets through Fandango or booking restaurant reservations through OpenTable. That’s the precursor pattern for a mobile personal assistant to consult and interact with a specialist.

Looking ahead in time, Assistants and Specialists co-exist, interact and evolve jointly and separately. Assistants started out as speech-enabled engines for natural language search and device control. Exemplified by Siri, GoogleNow and (soon) Cortana, they are highly-personal, context-aware keepers and interpreters of user activities, vital signs, preferences, entitlements, appointment calendars, notebooks and reminders.

Specialists benefit from deep understanding of domain specific information. If they are deeply integrated with corporate “back end” systems they have access to profile information (provided when you apply for credit or register a product), “entitlements” (based on membership in loyalty programs and purchase history), payment history, account balances and the like.  In addition enterprises obtain information from 3rd parties (like credit reporting agencies) which they use to decide how to treat each individual.

Banks, telcos, airlines, government agencies, insurance companies, health-care organizations and others put these specialists to work “to better serve you.” In the best case scenarios, they are very effective and speed authenticated users through their appointed rounds. Yet, by definition, they are not free agents programmed to operate strictly in the customer’s best interest. They are “owned” by the enterprise and must take into consideration such factors as the likelihood that a customer is attempting to commit fraud or that he or she does not qualify for the offer(s) that the company is extending to a targeted market segment.

This note just starts to addresses the challenges of promoting efficient assistant-to-specialist communications. Natural Language Understanding, Machine Learning, Artificial Intelligence, and Interaction Analytics will play a role in both in the foreground and background to smooth the way toward productive interactions between and among individual customers and the resources in an enterprise, live agents or advisors as well as Web-based chatbots, interactive voice response (IVR) units and search results. In the best case, smartphones (with the promise of consistency and continuity across multiple devices) will serve as starting point, compass and sense-maker of our personal journeys.

(Picture courtesy of Artificial Solutions)

Categories: Intelligent Assistants, Articles

Tags: Intelligent Assistance, Kasisto, Artificial Intelligence, Siri, Natural language processing