Artificial intelligence (AI) is everywhere — in our phones, in our homes, and in our customer service. It’s estimated that by 2023 there will be 8 billion voice assistants in use, and that’s just the beginning.
“The hype around AI is significant,” says Tom Hebner, head of product innovation, voice technology and AI, at Nuance Communications. “We’ve lived through the pain of the ‛marketing bot.’ They make a splash but people don’t use them because they don’t actually solve a problem.”

The smart future

Hebner has spent decades on the cutting edge of AI and voice user interface design, thinking about how close we are — or aren’t — to a “smart” future.

“The whole thing with conversational technology is, what makes things easier?” Hebner says. “Why isn't it everywhere? Because it's still hard. You still have to buy those certain pieces of technology. You have to integrate them yourself and tinker with them. Until all of that is in one ecosystem, you're still going to have some of those challenges.”

All about the data

Hebner and Nuance are at the forefront of breaking down those barriers in the name of improving both customer experiences; a recent study by Frost and Sullivan found that 77 percent of customers were moved to switch service after one frustrating interaction.

“We use AI technology to make lives easier,” he says. “Making interactions easier when people are interacting with banks and airlines, or improving the quality of communication when folks go to the doctor — and also making the lives of doctors easier by taking a lot of the manual work they have to do out of their hands and doing it automatically.”

The road to making lives easier is paved with data — and the knowledge of how to use it. “Natural language understanding performance has improved significantly over the last decade, but the NLU technology itself wasn’t what changed” Hebner notes. “The reason the performance got so much better was access to data — knowing what data to use and what data not to use.”

Blazing a trail

Hebner and Nuance turn that expertise in leveraging data into products that improve experiences — and thus return on investment (ROI). “We recently announced an innovation called Project Pathfinder,” he says, “leveraging unsupervised machine learning to ingest conversations (whether they are voice or text-based) and automatically create conversation paths. For example, everyone that's contacting a bank about activating a new card, we can see every conversation graphically, and see that your best agents have the conversation this way, while your newest agents have it this way. Supervisors can use that to craft training — here's how I can make my humans better. Even without that supervisor, we can recognize that this pattern is the most efficient, this is the least efficient, let's alert the newer agent in real time and give them the next best action so they can perform like the best agent. Our vision here is that we will make every agent your best agent.”

When it comes to fraud prevention, the ROI is easier to see but no less important. “With our biometric-based conversational AI technology, we can recognize people's voiceprints and behaviors — from how they talk to how they tap, type and text. That technology is almost magic these days because it’s allowing organizations to authenticate customers in a whole new way and in the meantime halt fraudsters trying to perpetrate attacks. By implementing biometrics, organizations not only improve the customer experience but save millions of dollars in fraud. That’s a perfect example of how we’re taking the potential of AI and applying it to a real problem and ultimately, when implemented, it drives impactful ROI for our customers.”

For Hebner, the future of AI and customer engagement goes hand-in-hand with making lives better. “The question is, are you making something they're doing easier?” he says. “Will it have value? Will it solve a problem? That's really where we live.”