When Watson played Jeopardy, one of the more powerful demonstrations of Artificial Intelligence was Watson’s ability to use Natural Language Processing (NLP) to field questions and respond with his answers. Behind the scenes, Watson was converting Alex Trebek’s questions from speech to text and from text into search arguments. When Watson then selected the most probable response to the question from the answer set resulting from that search, he was then able to convert that text into speech and use speech to present his responses. All this happened in less than a second, preventing Ken Jennings from being able to press his own response button in time for the majority of questions. The Jeopardy Contest became the launch pad for IBM’s Watson business unit and should rank as probably one of the more notable proof of concept projects in IT history.
Since the Jeopardy contest, NLP has gained significant traction as a communications pathway in business and in our personal lives. Powered by NLP, Apple’s Siri and more recently Amazon’s Alexa have emerged as models of a personal digital assistant (PDA). PDAs powered by NLP are now showing up as packaged software applications from the major independent software vendors (ISVs) that the consumer can configure and personalize on both the internet and mobile channels.
I watch my wife use NLP daily on her Samsung mobile device to compose her text messages to our children, our extended family, friends, and her business associates. As a home health nurse, she also uses NLP to enter unstructured text comments into her nursing visit reports. Companies in the hospitality industry are using NLP in text bots to greet guests after check-in and offer an alternative channel for order entry of room service and other hotel services like spa and golf course reservations. In several other industries, companies, including LaserShip, where we currently serve as CEO and CIO, respectively, are adopting chat bots as an alternative channel to voice for customer service delivery.
What are the common business drivers gradually promoting NLP to be the preferred UX/UI experience for today’s consumers and work force, or collectively our target customers?
• Emergence of the Mobile Channel–For today’s customers, the mobile channel is fast becoming the preferred channel of choice because it offers the time, location, and process convenience of the internet channel but extends this convenience to anywhere and anytime while the end user is on the move. Despite major advances in UX/UI design as demonstrated by companies like Uber and Lyft, however, the form factor of the mobile channel continues to present challenges to any application that requires more complex workflows and significant data entry. NLP can help reduce this complexity in applications requiring either personal or business communications.
• Safety–For any company in the transportation industry there is the ever present risk of distracted driver behavior. Traffic fatalities and accidents continue to climb and distracted driver behavior is a major contributing factor in the increase. National and most local government transportation jurisdictions have already passed regulations holding distracted drivers accountable for their behavior. NLP in the form of speech to text and text to speech has and will continue to make a major contribution to safety and compliance in helping to reduce distracted driver behavior, accidents, and fatalities. And even after autonomous vehicles and taxis hit critical mass and replace humans with robots as drivers, the human to machine communications that direct the autonomous vehicles will likely take the form of NLP.
• Customer Expectations and Preferences–Amazon and every other would be ecommerce competitor have trained today’s customers to adopt the following preferences, expectations, and behaviors. NLP can help meet these expectations in a variety of ways:
1. Customers want choice and control over their shopping, order entry, and customer service experience With options to use channels like text bots and chatbots as alternatives to contact centers and voice.
2. Customers want interactions with service providers that are personalized and relevant to their immediate context–Use of NLP channels and interfaces are offering customers the personalized experience of communicating with service providers in their language of choice. This feature appeals to a work force whose diversity is growing. NLP also offers the best way to communicate relevancy and context using the rich vocabulary of natural language.
3. Service providers must demonstrate authenticity, integrity, and speed to problem resolution via use of alternative channels to solve customer service problems–If the service provider demonstrates empathy and engagement with the customer and solves the problem quickly, honestly, fairly, and effectively, they are a hero. If not, they are toast. Either way, today’s customers tell everyone on the social network channels. Use of NLP channels to speed up effective service problem resolution can be instrumental in attracting and retaining customers and in helping to protect a company’s internet and social network reputations.
4. Ecommerce companies are training customers to expect instant gratification in solving their service problems–“First interaction resolution” is now the gold standard for service problem resolution. Lower latency and friction in customer service interactions can result from using NLP channels. Using NLP channels customers can communicate their orders and customer service problems quickly and get an immediate response back from a bot displaying search results, confirming their order, or solving their problem, versus sitting on hold on the voice channel, or attempting to solve their customer service problem using asynchronous communications like email.
5. Workforce empowerment and increased productivity Can result from the use of NLP as a low friction, “low code” user interface. An office supply company recently created a customer app that allows office managers to walk through the office and order supplies using an NLP interface on a mobile device. In conjunction with other forms of Artificial Intelligence such as Machine Learning (ML), NLP is enabling business analysts and data scientists to shrink the time to business insights and actionable information. Instead of asking IT to rewrite code, business analysts and data scientists can use NLP and speech to text as the method of defining and redefining the problem to the ML application. In doing so, NLP is transforming the way business analysts and data scientists work to solve today’s business problems and reducing the IT enhancement and maintenance debt of high value business intelligence applications.
Watson’s successful debut in the Jeopardy Contest demonstrated that NLP and AI based platforms like Watson could meet the expectations of today’s customers and workforce. Within the next few weeks, LaserShip is adding Watson Conversation Services and a Chatbot application to our application portfolio to supplement the services of our contact centers that already deliver high touch customer service on the voice channel to our shippers and consumers. We view this NLP application as instrumental in enabling our growth strategy and meeting customer expectations over the next several years. But we also believe that the Chatbot is only the beginning of our journey in learning how to use NLP and AI to meet the needs of our many other internal and external customers.