Page 27 - FCW, May/June 2018
P. 27

                                                   for example, who are overseas and may be interested in benefits, but
are in different time zones. What chatbots provide is that conversational interaction that can be highly personalized 24/7. They also provide the ability for agencies to go through peak periods or seasonal workflows, like Census or IRS, to keep up with
the almost impossible number of calls hitting their call centers.
Amalfi: Oracle in particular has been very active in this space. We’ve been working with the GSA and the Office of Emerging Technologies for about two years, alongside Microsoft, Facebook, Google, Amazon, and others. We’ve been working on pilot programs and participated in hackathons. There is a lot of interest, but they’re in the initial phases of starting to deploy this.
Q3. What aspects of Oracle’s solutions are best positioned to assist the government as it explores deploying this new technology?
Kuenzig: Oracle is approaching this market in a unique way. We’re providing a suite of tools to build chatbots across the enterprise and refine them over time, all in the cloud One of the core capabilities of the Oracle chatbot offering is to leverage AI and machine learning to enable end users to communicate with these technologies using natural language.
We are also empowering IT and business to work more collaboratively, creating chatbots using business- friendly tools like Microsoft Word and Excel. It’s important that the businesses that understand the language of
the organization can help shape the creation of these chatbot conversations. It also should be easy to rapidly update and deploy chatbot changes. We want to make sure these conversations are contextual, so we are enabling rapid integration through prebuilt adapters, which facilitates integration across the enterprise. Lastly, we are focused on building chatbots that help drive cross- channel consistency. This enterprise approach ensures that end users get the same information regardless if
they interact with a chatbot or more established channels like web and mobile.
Amalfi:There are many different types of questions someone can ask. One could be a general question that queues up an answer and you’re done. But there are also questions that are
a bit more complex, and the citizen wants to start doing transactions straight from the virtual assistant. At that point, it’s important to authenticate who the citizen claims to be. That’s where we deploy identity management capabilities, the same as those used on a portal. So we have built the solution to ensure that level of security is inherent to it.
Q4. What do you see as remaining challenges chatbots face for government implementation?
Kuenzig: One of the conversations that is evolving is this idea of trust and newer technologies. When users first try out a new technology, they have an expectation in mind as to what it can and can’t do. It’s vital when you release these products that there’s enough functionality to secure trust early -- and maintain it.
Amalfi: When you are responding
to a question as a government agency, you have to be confident the answer you give is consistent with the law. Agencies have been working with Twitter, Facebook, and other companies to review their terms of service to ensure they’re government friendly and address privacy issues. The government needs to make sure that when it deploys a new technology that affects millions of people, it’s not going to affect them negatively.
Agencies can also reduce some of the fear around using AI. People fear
AI will steal their jobs, their information, and so on. But once people see they can interact with government the same way they text colleagues, using the same application, and with reduced wait times and increased ease of use over sitting on hold or searching a website, citizens will eventually gravitate to leveraging and trusting the technology.
 SPONSORED BY
          
















































































   25   26   27   28   29