This Jan. 13, 2011 file photo provided by IBM shows the IBM computer system known as Watson at IBM's T.J. Watson research center in Yorktown Heights, N.Y.

A small Lawrence business is helping to harness artificial intelligence to improve the United States Air Force acquisitions process — a project the business can see through if it’s selected this spring to receive a second round of funding.

Lawrence-based KalScott Engineering Inc. was chosen last summer to train a supercomputer, or “cognitive thinking machine,” that could understand Defense Department contracting rules and answer questions about them. The Air Force initiated the project with the goal of streamlining a complex contracting process so it’s accessible to more businesses, making contracts — and prices — more competitive, according to an Air Force news release.

In the past several months, KalScott has fed tens of thousands of pages of contracts and contract regulations into IBM’s Watson, said Suman Saripalli, a co-owner of KalScott. Watson, a question-answering cognitive tool, has won “Jeopardy!” and has been used by medical professionals to answer clinical questions.

Jeopardy! contest Ken Jennings, who won a record 74 consecutive games, refers to his opponent, an IBM computer called Watson, while being interviewed after a practice round of the Jeopardy! quiz show in Yorktown Heights, N.Y., Thursday, Jan. 13, 2011. (AP Photo/Seth Wenig)

“We’ve taken all this data, and it’s been ingested into Watson,” Saripalli said. “It has to be trained for it to gain an understanding — what we call a ‘deep learning.”‘

A 2006 report from the federal government’s accountability office said the current federal acquisitions process “discourages small and innovative businesses from partnering with the government.” Camron Gorguinpour, director of transformational innovation for the Air Froce, said in a news release that cognitive tools, like what KalScott is working on, can help businesses “better navigate what is a very complex bureaucracy.”

Businesses applying for defense contracts, as well as government employees who work in procurement, can ask questions and get immediate responses from the system in a natural language. The computer is like a “very intelligent assistant,” Saripalli said, and would engage in a back-and-forth similar to knowledge navigators like Apple’s Siri or Amazon’s Alexa.

Saripalli and Tom Sherwood, both Kansas University graduates, opened their engineering and research and design firm in 2002. Drones are the company’s “bread and butter,” Saripalli said, and it’s previously been selected for jobs with NASA and the U.S. Navy to develop the unmanned air vehicles.

Saripalli and Sherwood are applying the same core technology the drones have to help train the supercomputer. Washington, D.C.-based Applied Research in Acoustics LLC was also chosen last summer to help with the project.

An Air Force news release says the tool should be completed by summer 2018. An announcement on whether KalScott is selected to finish the project may be made in the next couple of months.