Market Spotlight: Government
It’s no big national secret that governments on the federal, state, and local levels have been dragged kicking and screaming into the 21st century. In all of these sectors of government, the technology gap is massive, largely because of a lack of funding, knowledge, or a desire for change.
The tide is turning, though, and many more government agencies are starting to see the value in artificial intelligence–based applications that include speech technologies and their potential to help cut costs, free up millions of labor hours for more critical tasks, and deliver better, faster services.
This is most evident in the prevalence of automated 311 systems. Baltimore launched the first of these program in 1996 to help deflect non-emergency calls away from 911. Today, most major U.S. cities have adopted similar programs, and they have grown to be much more ambitious than the original 311 hotline. The programs have moved away from limited voice-based call center applications to incorporate other communications modalities, such as text, email, web chat, and social media. Some have even added automated machine translation technologies to accommodate citizens who speak languages other than English.
Further demonstrating the popularity and success of these programs, many other government agencies across the country have adopted similar technology for their constituents.
One example is the Emma virtual assistant launched in 2015 by the U.S. Department of Homeland Security’s Citizenship and Immigration Services. Emma (named after Emma Lazarus, the poet who penned the words that appear on the base of the Statue of Liberty) answers roughly half a million questions a month and helps visitors navigate the agency’s website. Emma is fluent in both English and Spanish, and though she always types out her answers in both languages, she can also speak her answers through a text-to-speech application that is part of the English version. Emma works on desktop and laptop computers and mobile devices.
Another example is the U.S. Army’s Sgt. Star, a voice-enabled virtual guide to assist visitors to the Army’s recruitment website. The Army launched Sgt. Star on its website in 2006 and added a Sgt. Star mobile app for Android and iOS smartphones in 2012.
According to the Army’s accounts, Sgt. Star, which was developed with Next IT (which was just acquired by Verint Systems in December), does the work of 55 recruiters, is accurate in more than 94 percent of interactions, and has increased the amount of time visitors spend on the site from four minutes to more than 10 minutes.
While 311, Emma, and Sgt. Star are relatively simple applications of voice technologies, developers are thinking bigger, looking for ways to expand their penetration into the public sector. Speech analytics is one area that is showing real promise.
Verint Systems recently rolled out Verint Speech Analytics in one of the Hong Kong government’s Efficiency Unit contact centers to uncover citizen insights, trends, and behaviors.
The contact center handles inquiries and complaints from the public. Currently, it has 22 participating government departments. Each year, it receives about 4.1 million calls.
The advanced functionality in Verint Speech Analytics will help provide the agency with semantic intelligence to automatically identify and organize the words and phrases spoken during calls into themes that can then be used to spot service deliverability problems and other common issues.
“By helping us be more productive in prioritizing inquiries and performing root cause analysis, we’ll be able to more easily determine emerging trends through the insights gathered. This will not only allow us to address our objective of reducing the need for the public to go through the trouble of making inquiries, but also enable us to improve public service delivery,” said Simon Lam, a principal consultant for the contact center, at the time.
But government agencies across the board are exploring a wider variety of uses for speech technology. Among them are speech recognition, which has obvious benefits for transcription, dictation, and phone assistance; machine translation, which has obvious uses in international relations, defense, and intelligence; and voice biometrics, which has wide government implications for security, intelligence, fraud prevention, and more.
Even text-to-speech technology has found its way into the public sector, with a number of transportation agencies, for example, using TTS to inform passengers of upcoming stops, service alerts, notifications, reminders, and other information.
To show the importance of this sector, Acapela Group has dedicated an entire business unit to transportation. Some of the government entities that have relied on solutions from the Acapela Transport unit include Jernbaneverket, the Norwegian government agency for railway services.
“Pleasant voices are now deployed to inform our passengers with highly accurate audio restitution, resulting in, we hope, a pleasant traveling experience,” said Victor Hansen, head of customer and traffic information at Jernbaneverket, when the service was first launched.
Similar technology is being used for a virtual assistant at Japan’s Narita Airport and at kiosks for buying bus tickets in Sweden and Norway.
Another strong market for TTS in the government sector is in making websites accessible to the disabled. Through assistive technologies that incorporate TTS, government websites can feature voice capabilities that can read aloud site content and respond to voice commands to navigate.
AudioEye, one provider of this kind of technology, recently partnered with CivicPlus to expand its services to the government.
“Digital accessibility is a key initiative for CivicPlus on so many levels. It is important not only for the sake of compliance with regulations, but more importantly, to ensure all citizens have equal access to the information made available to them from their local governments,” said Sascha Ohler, CivicPlus’s vice president of research and development, in a statement at the time.
One of the first government agencies to adopt the AudioEye technology was the U.S. Social Security Administration, which in October chose AudioEye to provide web usability tools for the socialsecurity.gov website. The site draws between 12.9 million and 22 million visitors every month.
“Nearly nine out of 10 individuals age 65 and older receive Social Security benefits. Given that 64 percent of the U.S. population 65 and over are active Internet users and more than 6 percent have a visual disability makes a strong single case for the need to improve usability of the SSA site,” said AudioEye CEO Todd Bankofier in a statement at the time. “However, that is only one demographic for which our tool will enhance the experience. Many levels of cognitive disabilities can be addressed as well as those that impact mobility, hearing, or vision. This is a great step towards even greater inclusivity.”
And as advances in speech and artificial intelligence technologies continue, many expect massive changes in the private sector, transforming how government entities interact with their constituents. It’s likely to eliminate some jobs, lead to the redesign of others, and create entirely new professions. And though adoption has been slower than in the public sector thus far, the technologies will fundamentally change how government works, and those changes will come much sooner than many think.