Table of Contents
The focus of this Market Insight is the use of speech or text-driven agents as a supplement to or replacement for live agents in the contact center. Interactive Voice Response systems (IVR) and other self-service channels have been used to offload more expensive agents in the contact center for decades. More recently, Virtual Agents – automated, speech and text-enabled – have introduced a new level of self-service capabilities to the contact center. This insight explores the dynamics surrounding the use of these agents to offload and assist live agents in the contact center, and where things stand in this realm heading into 2014.
Despite considerable hype, vendors with speech-driven solutions that were early entrants to the market back in the mid-1990s didn't gain much traction. Indeed, a large number of those early companies—General Magic, HeyAnita, Native Minds, Webley, Wildfire, and others— were either acquired or disappeared. Back then, there seemed to be a perception that speech technology applications, including Personal Assistants and natural language IVR applications, weren't robust enough, and so wide scale adoption of Speech Assistants was stymied.
But the situation today is different as Virtual Assistants in the contact center are fast becoming commonplace. This Market Insight examines the history and landscape leading up to this development, the technology advancements fueling this change, and the solutions enterprises should investigate in determining if Virtual Agents are a good choice for their contact centers.
Making Sense of the Automated Landscape
The term "Virtual Assistant" has been around for a long time, but over the years other terms have entered the lexicon, including "Automated Assistant" and "Personal Assistant". All these "agents" or "assistants" use speech technologies at the core and provide varying levels of functionality. As such, they are not solely focused on live agent replacement. There is a lot of overlap in how they are used, but even though the terms are often used interchangeably, it is worth better defining the different labels.
The concept of a speech technology-driven Personal Assistant first surfaced in the early to mid-nineties. By 1996 there were over forty companies that had developed products in this space. These solutions were launched primarily to provide access to office productivity functions and information of general consumer interest. For instance, companies such as HeyAnita!, Webley, Wildfire, and General Magic allowed users voice-enabled access to calendars or telephony functions, such as conference calling and placing calls. They also provided access to a set of information "sites" such as weather, sports scores, horoscopes, stock quotes, and even soap opera updates.
On the consumer side, these Personal Assistant solutions were the precursor to today's SIRI—Apple's Personal Assistant on the iPhone. However, their scope was limited, and monetization by solution providers focused on the application rather than making the phone itself more useful.
Today's Personal Assistants—now more commonly termed Virtual Assistants—are speech-driven applications designed to assist consumers in doing things such as accessing information on mobile devices. Virtual Assistants take advantage of the broad capabilities provided on smartphones, making the phones themselves more useful in hands-free mode. A well known example, Siri, is a consumer grade Virtual Assistant that makes it easier to use a mobile device. "She" allows users to do numerous tasks including:
- Search the Web
- Ask for directions
- Check weather, stock prices, restaurant reviews, etc.
- Set reminders and take notes
- Play music
- Place phone calls
- Launch photos
- Send message
- Post to social networking sites
- Chat Bots
Chat Bots, also termed Chatterbots or Chatter Robots, are artificial intelligence (AI)-based computer programs that simulate an intelligent conversation with a person through text or speech. Using natural language processing (NLP) and knowledge databases, they are typically used to "chat" or engage in small talk with the user. They can be presented as an avatar, which is an engaging graphical representation of the Chat Bot's alter ego or character. Chat Bots can be used as companions in software programs like robots. They also can be used to chat in software programs or online in the same way that virtual chat agents are used.
Virtual Contact Center Agents - Speech and Text
Virtual Agents are conversational agents that apply AI and machine learning to customer self-service. These agents dynamically generate dialogue with the customer and personalize the interaction. Virtual Agents have "short-term memory" in that they can apply context within a conversation, and "long-term memory", pulling from and adding to customer interaction data records. Virtual Agents can have persistent conversations by "remembering" where they left off with the last interaction. They also can be self-learning, adding to knowledge repositories so that responses get better over time. Although initial solution provider products were focused on front-ending a contact center, text-driven chat agents are also increasing in prevalence. The Avaya Virtual Agent (Ava) and Creative Virtual's V-Person products are two such examples. They use NLP and knowledge databases to create chat conversations with customers. Consider that Ava accesses a knowledge base built on pre-defined question/answer pairs of potential customer inquiries.
However, as Ava handles more and more chats, "she" can "learn from" live agent handling of questions and add new responses to the database. This is evidenced by Avaya's own use of Ava on their customer support portal. Avaya support engineers are required to add solution content within 90 minutes of closing a service request, and as part of this, they develop new question and answer pairs for Ava to access.
Solution Provider Approaches and Use Cases
Apollo Enterprise Solutions (AES)
Apollo Enterprise Solutions (AES) TRUE Agent Emulation System for Credit Portfolio Management is an example of the breadth of work that can be done using Virtual Agents. It is an enterprise-wide Software-as-a-Service (SaaS) solution consisting of multiple customercentric modules that allow financial organizations to manage the entire credit portfolio lifecycle using self-service Virtual Agents. Powered by AES' proprietary Intelligent Decision Engine, which automates the decisioning of complex financial transactions based on business rules, the system presents to customers in real-time individually tailored choices based on the customers' own financial situation.
AES's solution differentiates itself through its ability to take the individual's financial profile, including information from multiple creditors as well as credit bureaus, and generate realtime choices and possible settlement offers for the customer. The system provides customers who are in financial distress with choices for resolving debts, and educates customers on the effect each choice they make has on their creditworthiness. This enables them to immediately see how various possible scenarios will impact them, allowing them to make better choices. Indeed, the system enables the customer to self-serve at any time, through any device, from PCs and mobile devices to ATM machines. Unlike typical IVRs, the system also allows the customer to save the work they have done and come back and finish later.
Founded in 2004, Creative Virtual's vision is to replicate a contact center online without the cost of live chat. To fulfill on this idea, the company created a natural language system integrated to back-end systems, providing specific answers to a customer. As both text and speech-driven agents, V-Portal and V-Person Virtual Agents are omni-channel in the sense that they can be used with IVR, mobile, tablet, social media, and Web customer interactions. For instance, in a mobile application, the agents use speech recognition or text, have support for HTML5, can provide personalized display information on the user's device, and access knowledge databases through V-Portal.
In May, 2013, IBM announced that it is applying its high-profile Watson technology to customer engagement through the first broadly available commercialized offering built on that technology. The new solution, called IBM Watson Engagement Advisor, can be used as a Virtual Agent in customer service engagements. Thus far, Watson technology has been used in areas such as healthcare and finance. For instance, using natural language capabilities, hypothesis generation, and evidence-based learning, Watson helps physicians cull through mass amounts of data, assisting them in diagnosing and treating patients.
Watson is a cognitive platform that can:
- Process vast amounts of data far more efficiently than a human agent
- Understand natural language
- Place content in context for greater insights
- Generate and evaluate evidence-based hypothesis and recommend responses
- Learn and adapt to ever-changing flows of new information
Get Industry Insights. Simply.
Talk to Veronica
+1 718 514 2762