Speech Technology in the Field: Case Studies in Speech Recognition, AI, IVR, Chatbots, and more
As part of this year’s Speech Industry Awards issue, we wanted to give some attention to the many innovative uses of speech technology—and, more broadly, artificial intelligence (AI)—we’ve seen throughout the year in almost every industry vertical. Read on to find out more about just a few of the implementations that caught our eye in 2018.
Making Sense of Cryptocurrency
Cryptocurrencies are all the rage today. Industry analysts lump blockchain and bitcoin in with artificial intelligence, machine learning, and the Internet of Things as emerging technologies with the potential to disrupt many well-established industries. Like any new technology, many questions arise about cryptocurrency. In fact, many individuals do not know that blockchain is an underlying computing technology and Bitcoin is a virtual currency. Others do not understand how cryptocurrencies, like Bitcoin, differs from paper money or credit cards.
Enter Ben, a chatbot that acts as an information portal. The software, which runs on Apple iPhone and Google Android mobile systems, enables users to take classes about cryptocurrency, read the latest industry news, and even buy and sell Bitcoin tokens. After completing a KYC check (which authenticates an individual’s identity), consumers in 21 U.S. states can buy and sell Bitcoin tokens. The start-up charges 1% for a buy or a sell, which is similar to non-virtual exchanges. Customers can then track their Bitcoins’ performance on their phones.
Ben’s founders have raised $580,000 in pre-seed money from angel investors Third Kind Venture and Y Combinator, which runs a Shark Tank-like competition and awards winners with money and early-stage business development assistance from Silicon Valley executives.
Support for other blockchain technologies, like Ethereum, Ripple, and Bitcoin Cash, are under development. The blockchain market is expected to grow from $411.5 million in 2017 to nearly $7.7 billion in 2022 (a compound annual growth rate of 79.6%), according to research firm MarketsandMarkets. Ben plans to provide information to individuals as this market takes shape.
Target Takes Aim
How companies market to customers is in constant flux. As voice has gained more acceptance from the use of smartphones and home automation systems like Google Assistant, retailers have been searching for new ways to leverage the technology. Target, a retailer that generated $71.8 billion in 2017, tried out one speech application that looks like a springboard to additional revenue.
In the spring of 2018, the retailer teamed with Google to develop a voice-based marketing program relying on many different components. One of them, Google Express, is the tech giant’s home delivery shopping service. After testing the technology in California, Target became a Google Express partner and rolled out the service to U.S. customers in October 2017.
Shopping Actions, a Google Express add-on, works with Google Assistant to allow consumers to add products to a universal basket that connects mobile, browser, and voice platforms for ordering. Target offered users a $15 coupon toward their next order to encourage them to use the service. Consumers activated the discount by saying or typing “spring into Target” on their Google Assistant app on iOS or Android, or other mobile devices using the voice activation system.
Results from the Shopping Actions trial were promising. The promotion was originally slated to run a month but was cut short by more than two weeks because its goals were reached. Target found that the number of items in customers’ Google Express carts jumped by an average of $20. Because of the success, the retailer is now looking to build upon the success with another similar promotion.
Cricket Gets Chatty
With more than 2.5 billion fans worldwide, Cricket is the world’s second most popular sport. Tracking the results from matches around the world has been a challenge, one that Roanuz, an Indian AI speech recognition start-up, has taken on.
The firm developed a chatbot that answers natural language queries in both text and voice. The system mines answers from important tournaments and addresses questions about historical matches, player information, and team statistics. The leagues tracked include the Indian Premier League, Big Bash League, Caribbean Premier League, Women’s T20I Tri-Series, Kenyan Premier League, Champions League T20, and more.
In addition, the system provides fantasy league players with information about how their teams are faring. Consequently, the solution is designed to work in near real time. The chatbot has processed more than 1.2 million calls. The system currently responds in about 1,500 milliseconds, and the vendor is trying to lower that number to 300 milliseconds.
Many fans thirst for information about their favorite sport. Now, Roanuz is providing them with a way to access it more easily.
Data Entry Gets Easier
Healthcare providers often struggle trying to improve patient care while holding down costs. Japanese healthcare technology provider CMIC Holdings has built a voice recognition system designed to speed up the inputting of clinical trial information. The solution, which runs on iPhones, is based on AI voice recognition technology from NEC Corp. The vendor has a growing suite of artificial intelligence solutions, marketed under the moniker of NEC the Wise.
Traditionally, healthcare companies entered clinical information manually and clinical research coordinators (CRCs) checked the data for errors. With the CMIC application, healthcare personnel enter data required for drug clinical trials—such as a subject’s height, weight, medicines taken, and lab test results—verbally.
The change reduces the errors and omissions that occur whenever information is entered by hand. For instance, if healthcare professionals miss a required field, the system will not let them move to the next step in the process.
Security is a key element in the medical community. The CMIC application issues an audio warning if a data value outside an expected range is entered. A pilot program began in March 2018, with more than 50 clinical research coordinators as part of CMIC’s Site Management Organization program. The plan is to expand the program in the coming months.
Putting the Customer First
Faced with rising consumer expectations, Royal Bank of Canada (RBC)’s contact centers, which serve millions of customers per year, felt hamstrung by their legacy system’s design. Customers of North America’s sixth largest bank often got lost fumbling through the touch-tone interactive menus and abandoned their calls.
To better serve its clients, the financial institution decided to replace the system in 2016. The bank wanted to boost automation, minimize adviser-to-adviser transfers, and improve customer satisfaction with conversational dialogues. After evaluating potential replacements, the bank opted for Omilia’s Conversational Virtual Assistant Banking.
Once the system was deployed, agents became more productive. The solution supports omnichannel interactions: interactive voice response (IVR); chat; social messaging platforms (Facebook Messenger, Twitter, and more); and even Internet of Things home speakers, like Amazon’s Alexa and Google Home.
The AI voice recognition system supports 17 languages and features adapted acoustic and language models that account for accent and dialect variations within a country.
Making the switch led to many business improvements:
- Double-digit increase in containment within six months of deployment
- Increased client satisfaction
- Discovery of new convenient services to offer clients
Many established corporations have relied on legacy IVR systems that have grown long in the tooth. With technology advancing at a lightning clip, perhaps more organizations need to make a switch à la RBC.
We present the thinkers and innovators who are creating new tools and approaches for speech technology—and fostering the next generation of talent. In this installment, we talk to Ilana Shalowitz from Wolters Kluwer Health.
A Game Platform Helps Nonverbal Children Find Their Voice.
These cutting-edge vendors are leading the way in AI, analytics, natural language, smart speakers, and more.
Companies and Suppliers Mentioned