Artificial intelligence can absolutely help in the data challenge, but standards are key

One common thread ties the health care industry with the federal government, in that both have what feels like an endless amount of data to work with and much of it tied to specific people.

Now think of the Veterans Affairs Department's mission as a provider of health care and other benefits to those who served in the armed forces and their families.

During WT's April 12 Power Breakfast focused on VA as a customer, a panel of industry executives from the department's largest contractors highlighted the role of technology in helping VA in its mission and everything else needed on top of the tech.

After the PACT Act became law in August 2022, VA saw an influx in applications for claims due to their exposure to toxic burn pits and other similar pollutants that harmed their health.

Travis Burd, senior vice president and leader of Booz Allen's VA business, pointed out that each veteran can have up to 1,300 documents associated with their name. That can include service records, medical records, handwritten doctor's notes and other information on a veteran's health history.

"The ability to process the volume of claims quickly enough to get the benefits to veterans in the time they needed to be, hiring more representatives alone wouldn't be a sufficient approach," Burd added.

Booz Allen identified the PACT Act exams and evaluations as a "great use case" to apply artificial intelligence and other automation techniques to work with all that data, Burd said. The idea was to help claims agents make decisions and quickly go through each veteran's file of 1,300-plus documents through the use of AI.

What else is needed though besides the tech?

"A team that understands the processes – be it the benefits life cycle, be it the clinical setting, which is why it's so important to partner with veteran-owned businesses," Burd said.

Tim Gilday, senior director of emerging technology at General Dynamics IT, zoomed the discussion out with an analogy to trying to find something in the deep seas.

"Any government organization is sitting on a treasure trove of data, and I think anybody that has had to deal with that realizes it's like a treasure box that's at the bottom of the ocean, and you need to build a submarine that can withstand that depth to get at the riches there," Gilday said. "It's just as difficult to actually take advantage of some of this data because it's not clean, it's not easy to match."

The labor-intensive effort to clean up that data so it can be used and analyzed does help to gain insights and make decisions. Gilday said AI will only increase the amount of cleaned-up data and analysis after the fact, plus be used to push other applications that can rest on top of that data.

An organization like VA with the mission it has must deal with the kind of increase in data coming in when laws such as the PACT Act create a massive influx.

Then there is the aspect of standards, or in essence how data can be handled by one team inside government versus another.

"We are getting to the point where AI can help us to clean that up, but we'd still rather we have those standards in place to make it a lot easier between today and that day when AI can really accelerate that process," Gilday said.