Sarah is an agent working in a call center on a call with a customer.
“What seems to be the problem?” she asks.
"I am having a problem with my keyboard,” says the caller.
“Can you be more specific?” she inquires.
"Yes,” the caller says. “One of the keys seems to be stuck. I have to press it twice to work."
“What is the model number of your phone?” Sarah asks.
“Universal S17," the caller responds.
Suddenly, a list of items appears on Sarah’s computer screen. These items all pertain to sticky keypads on the Universal S17. Sarah scans the list.
“Ah I see,” she says. “It looks like we will need to schedule a return for you.”
The screen changes to a form for submitting a return. Much of the form has already been filled out with the client name and problem statement. As the customer gives Sarah his information, the form on the screen immediately populates the form, line by line, without Sarah having to type.
“Ok. We are all set,” she says. “How would you like follow up information delivered?”
“SMS, please,” he answers.
“Done!” she concludes. “Have a great day!”
This is not science fiction. In fact, the future is closer than you think. Artificial intelligence is powerful enough to listen and act on a conversation in real time, opening up an entirely new category of capabilities for the business.
Welcome to the world of conversational AI.
Conversations – The Next Big Thing
Conversations are the glue that hold our entire society together and are absolutely essential for conducting business, so it was only a matter of time before the IT industry would turn its sights on spoken dialogue.
Of course, we all know about the amazing buzz products like Amazon Alexa are creating in the market. And, just recently Microsoft (News - Alert) announced that it is forming a 5,000-person division to focus on conversational technologies. In addition, there is also a budding conversational commerce movement, and several new approaches are being introduced into the market to solve a wide range of use cases.
Many of these initiatives are focused on chat. Where they are focused on voice like Alexa, it is the fundamental conversation between man and machine. However, conversational AI analyzes and operates on voice conversations between two or more people. This elevates the complexity of the AI, but also makes things more useful. Even with all the advances in digital communications, voice is still king and can account for 70 percent or more of customer communications.
With two or more people talking, the technology has to be able to identify the different participants and extract meaning and context from all parties in real time. This wasn’t even possible five years ago, but thanks to advances in AI, as well as cloud services like Amazon and big data technologies, it is now possible.
AI’s Killer App
The big turning point started around 2011 with the emergence of a branch of machine learning called deep learning. Deep learning algorithms mimic the way the brain operates and can learn about their environments in more complex ways. Deep learning applied to speech recognition resulted in a quantum leap in word recognition and made possible speech tools such as Amazon Alexa and Apple (News - Alert) Siri.
Yet, it’s not just about the algorithms. Configuring how to scale as well as tune the algorithms for specific languages are equally paramount. Together, they present an enormous amount of complexity.
Here’s a list of some potential applications.
Conversational Mining: Being able to process many conversations and aggregate and summarize what was said. This kind of thing would allow product marketing managers to learn about the top objections from the last 1,000 sales calls. Or, a CMO could leverage it to quickly determine how her $5 million advertising campaign is going based on whether people are calling in and requesting the product. Undoubtedly, there is tremendous value when people know what is happening regarding sales calls and the like.
Co-Pilots and Form Fills: As we saw with the example above, listening to conversations and mining knowledgebases to find and present relevant information or extract information from the conversation and subsequently entering it into CRM applications presents another opportunity relative to conversational AI.
Being able to give businesses the ability to capture, analyze, and operate on their conversations to accelerate business outcomes, whether that be faster time to resolution on a customer call, or faster time to closing a deal, delivers game-changing capabilities. That could open the door for establishing speech (and, free text) as the new interface for communicating with machines.
Mohamed Afshar is president and CEO at Spoken Communications.
Edited by Alicia Young
