Using AI and data to make decisions faster

Federal IT workers have become increasingly aware of the importance of using data and artificial intelligence to meet missions, as well as the challenges  in doing so.

To get a sense of where agencies are with AI and data -- and where they still need to go -- GCN spoke with former Federal CIO Suzette Kent and Mark Herrington, CEO of OnSolve, a critical event management provider, about government’s use of AI and data.

This interview has been edited for length and clarity.

What is the current federal landscape in terms of AI and data?

Kent: When I was on the government side, I spent a lot of time talking about data, particularly making the data publicly available and usable. On the AI side, as you’ve seen over the past few years, the focus has been on not just research and some of our high-performance compute capabilities, but on applying the capability. The ability to take data from disparate sources, smooth out the processes and help organizations -- whether it’s a government entity or it’s a business -- make decisions faster in a crisis situation is a fantastic example of the evolution of both the technology and the use of the data.

Herrington: Right now, we’re seeing overwhelming demand for threat and risk intelligence from clients. The goal is speed and relevance of information. If you think about trying to avert or manage through a crisis, you’ve got to get information as quickly as possible -- every second and minute counts -- and that information has to be very, very accurate and relevant to the situation. We take in over 17,000 data feeds on a daily basis and identify 300,000 events that are happening around the world. You think about that volume of data and doing that quickly and with relevance, it’s just not possible to do that with human analysts.

What can agencies do to make the most of their data?

Kent: I’m going to give a nod to the work that is still continuing under the Federal Data Strategy. Agencies have been doing a lot to make data usable and to make it available and reliable. One of the things I’m excited about is places where federal, state and local can work together better. Those are places where there’s opportunity to push harder on the quality of what’s available and how it’s made available.

What challenges do agencies face in using these technologies?

Kent: I think there were some realizations during the pandemic -- this significant distributed work environment -- that every location is not the same. Agencies need additional precise information about that specific jurisdiction, whether it’s the employees or whether it’s the constituencies that they are serving. They also need communication that reflects that level of precision. AI and data help you manage both your overall enterprise but also the details associated with managing a distributed environment.

Herrington: It’s a matter of having technology that is current and modern. It seems like different types of crises are coming at us much more frequently now, so having sound business processes and practices are key. It’s also important to have platforms that are intuitive, easy to use and don’t rely on human analysis or reactions. This is where machine learning and artificial intelligence can act as wingmen.

There are many technology lessons to be learned from the pandemic. What are some takeaways?

Kent: As we had to move very quickly -- do things in days and hours, not months and years -- many agencies relied on vendor partners that they trusted and solutions that were new, modern and scalable. But there were many places where digital services were pushed out on the front end, but there are still inefficiencies in the backend. So those are opportunities.

There are other places where we realized maybe somebody had a piece of information or two but not the full spectrum of data. If they had a broader data spectrum, and it was moving through a faster engine, they could make better decisions. Breakthroughs in the vaccine development, for example, were absolutely attributed to AI and its ability to aggregate so much information and run so many variables at one time. As agencies continue forward, they need to take some of those lessons and – rather than make them into a special initiative --  embedded into the processes that they do every day.

Herrington: This country is great at getting better and learning from crises. You’re seeing it in the focus on continuity and resiliency both in the public sector and the commercial sector. This is now not No. 8 or 9 in their top 10 priorities; it’s No. 2 or 3 or 4. I think this black swan event will lead to a lot of improvement in terms of business process, leveraging data and AI, sharing information and being able to operate with speed and accuracy of information.

What can agencies do to take advantage of AI?

Kent: One of the things I said frequently when I was on the government side was just get started. Start on the journey with AI because that does a couple things. That forces you into discovery, both about your data, about the resources you have on your team, about how you establish those channels. This is something we must do. It’s not a special project. It is about survival and it’s about keeping your constituents safe, your employees safe, your businesses safe. This is a place where we can leap ahead with modern tools vs. other places where we’re dragged down by so much legacy technology.