KISSIMMEE, Fla.—Contracting shops at the National Geospatial-Intelligence Agency (NGA) are shorthanded, just as they are across the federal government, but artificial intelligence technologies exist that can make the agency’s acquisition processes more efficient and faster, NGA’s new associate director for capabilities, said this week.

Natural language processing, a form of machine learning technology, could be used to determine the best contracting approaches and develop lessons learned from past contracting, Trey Treadwell, who is also NGA’s component acquisition executive, said on Wednesday during the annual GEOINT 2024 symposium.

Right now, the agency is not taking advantage of these capabilities, Treadwell said during a fireside chat on the exhibit floor.

NGA already has an enormous amount of data about contracts that have performed well and those that have not and a bot would help program managers, contracting officers, and contracting officer technical representatives sort out the best contracting approach, he said.

“How do we take advantage of those lessons learned that isn’t inside a single contracting officer or a small teams’ lessons, and we incorporate it into an environment where we could all benefit from that knowledge and so that a small team of contracting officers can do the work of multiple orders of magnitude of what they would be able to do yesterday, but tomorrow,” Treadwell said.

Adopting AI capabilities within NGA’s business environment will require training and education for the contracting officers, and “top cover” from agency leaders to back their contracting officials in taking risks associated with the new technologies, he said.

“We know we want to lean forward and the opportunity cost of not doing that is greater than the actual cost of doing it,” he said.

Treadwell has been with NGA about four months, joining the agency from the Treasury Department. On LinkedIn, he bills himself as a “data evangelist, non-conformist, advisor, integrator, intelligent change agent.”