"The robots aren't coming, they are here," is all I could think of as I mulled over my notes from a recent meeting with Gregory Pal (News - Alert), Vice President, Marketing, Strategy & Business Development at Nuance Communications. Coincidently I also happened upon an article at the same time about how robots are doing more of the writing of the stories you read from the Associated Press – saving time and money and, more importantly, allowing sharing of stories that are of interest to various groups, but which couldn’t previously be covered profitably.
To me, it seems the technology used to write stories is similar, in some ways, to the natural language understanding, or NLU technology, Nuance (News - Alert) has been working on for many years. In fact, the company handles more than 14 billion customer engagements per year with 80 percent of its on-demand customers taking advantage of NLU, as well as 1.7 billion mobile and web conversations processed through virtual assistant solutions.
According to Greg, the company has just dramatically improved its NLU-based product line by adding the following:
Designed to extend the benefits of self-service automation to complex scenarios, Nuance has expanded capabilities to enhance the capture of email addresses, alpha-numeric case number and multi-part forms by allowing human agents to silently assist. The benefits include increased containment, reduced misrouting and an overall improvement in customer satisfaction.
Two-Way SMS with NLU Automation
Building on the growing popularity of text messaging for customer service, Nuance has enhanced its real-time two-way capabilities with Natural Language Understanding to remove the constraints of traditional text dialog. As a result, consumers can respond naturally and be understood by automated self-service systems – increasing containment, deflecting inbound calls and improving customer experience.
Nuance Experience Studio
Working in tandem with a suite of common platform tools, Nuance Experience Studio extends the ability to rapidly optimize Natural Language Understanding grammars to the Enterprise, leveraging investment in one channel across other channels, thereby lowering the Total Cost of Ownership associated with intelligent self-service systems.
Greg likens these innovations to artificial intelligence or AI for customer service. He says other systems focus more on keyword spotting and discrete commands like yes, no or stop which make them less useful.
If a prospective customer is unsure of how the system will work, they are welcome to force-feed it with input to gauge response. Moreover, there is the ability to add personality based upon the brand. Domino's, for example, names its virtual assistant Dom and during an order it might say, “Let’s take this to another level and add lava cake.” Or, if you ask it to marry you it will suggest you order pizza for the reception.
I decided to try Dom out for myself. I had never used the service so I called the local Domino's near TMC (News - Alert) HQ in Connecticut thinking it might be an IVR solution. The person who answered said I needed the app if I wanted Dom and that the app now accounts for half of the store’s orders. He said people love it.
When trying it – the accuracy was generally good – when I tried to trick it by ordering something like fries – it kept adding pizzas – which I suppose is a bit of a bug but then again, I was trying to trick it.
Getting back to Nuance – we discussed using this technology for social but, as the results are potentially seen publically, it may not be something most customers will choose to do at the moment. Over time, that may change.
For companies investing in omnichannel customer service – the power of AI-based NLU solutions will now follow them from the call center to the contact center – to the app and beyond. Let’s also hope more companies continue following the Domino's lead and invest in humanizing their AI/NLU personalities in order to make the experience as pleasant as possible for customers.
Edited by Maurice Nagle