Why the Pentagon’s JEDI Cloud Contract Wasn’t a Total Loss
Experts say the now-canceled project is a reminder that tech acquisitions need flexibility to take into account innovation.
Defense Department officials ended its long, strange attempt to buy a single enterprise cloud by canceling the Joint Enterprise Defense Infrastructure contract July 6 after years of legal challenges, Congressional inquiries and other delays that kept the project from ever getting off the ground.
But it surprised few in the government contracting community that JEDI’s days were numbered. Even when JEDI was conceived in 2017, few enterprises were locking into single cloud solutions, and some experts argue that JEDI was dead—or close to dead—on arrival.
“It struck me that at the time … that maybe this … was overcome by events,” Stan Soloway, chief executive of Celero Strategies and a former DOD acquisition official, said in an interview with Nextgov.
The JEDI concept was the right idea at the time, DOD acting Chief Information Officer John Sherman said after announcing the contract cancellation. But the department’s changing needs would still merit something like JEDI’s replacement, the Joint Warfighter Cloud Capability. DOD unveiled the JWCC concurrently with the cancellation of JEDI.
“Even if [JEDI] popped out of litigation last week, we’d still need this new approach embodied in JWCC,” he said. JWCC will be a multi-cloud, multibillion-dollar project. DOD plans to send direct solicitations to cloud service providers; it initially intends to solicit proposals from Microsoft and Amazon Web Services, though it may include others depending on the results of market research the department will conduct over the next few months.
Still, Soloway and other experts said that JEDI and all of its twists and turns may not prove to be a complete waste of time: The project may have encouraged components across the department to pursue cloud more aggressively as well as highlighted the need for acquisitions to be able to take into account the dynamism of the tech sector as it works to integrate emerging technologies into the agency.
Cloud growth at DOD
During the unveiling of JWCC Tuesday, Sherman cited increased cloud conversancy as an important factor contributing to DOD’s readiness for a multi-cloud environment. Between the rollout of JEDI and its cancellation, Sherman said cloud conversancy has picked up at the Pentagon.
Projects like the Air Force’s Cloud One and the Defense Information Systems Agency’s milCloud 2.0 have ramped up in the intervening years. The Army turned its cloud management office into the Enterprise Cloud Management Agency earlier this year, which Army Chief Information Officer Raj Iyer said represents a commitment to cloud acceleration. According to an analysis by Deltek’s Alex Rossino, DOD awarded nearly $400 million more in cloud contracts during fiscal year 2020 than it did in fiscal year 2019—and that’s excluding the JEDI award.
Experts told Nextgov that JEDI may have indirectly contributed to cloud growth at DOD. At a minimum, it brought conversations around the need to move to the cloud to the forefront at the department, they said.
Four years ago, many of these aforementioned cloud efforts were prototypical or experimental, Mark Testoni, chief executive officer of SAP National Security Services, told Nextgov in an interview.
“[JEDI] became an impetus,” Testoni said.
Michael Acton, vice president of solutions and engineering at Array Information Technology, characterized JEDI as a green light to go ahead and move to commercial clouds. Once the signal from the top came down, DOD components were comfortable exploring whether off-premise commercial clouds might be better options than on-premise offerings like milCloud, he added.
“There were a lot of regulatory and authorization problems trying to move to cloud before JEDI hit the street,” Acton said in an interview with Nextgov, “because people are like, can we move to the cloud? Can we get out of DISA? We don’t know.”
Chris Hughes, a consultant who has worked extensively on DOD cloud efforts, told Nextgov in an interview JEDI wasn’t the only stimulator for cloud movement at DOD, but the project did advance the conversation around cloud.
“People were still moving forward with utilizing cloud and maturing in cloud despite the absence of JEDI,” Hughes said. “But I definitely think that JEDI did contribute to the conversation in bringing some of these topics to the forefront.”
Moving forward, DOD doesn’t want its components to abandon the fit-for-purpose cloud options it’s developed. Sherman said the department is not “pulling the throttle back” on programs like Cloud One or milCloud 2.0. General Dynamics Information Technology, the contractor that supports milCloud 2.0, said in a statement that milCloud 2.0 is “ready to meet the DoD’s critical and urgent enterprise cloud requirements, providing rapid access to both secure on-premise and general purpose cloud services.”
But none of these other options fill the unmet need for an enterprise cloud that can support Joint All-Domain Command and Control, or JADC2, which is the Pentagon’s connect-everything concept meant to support joint warfighting, or the recently announced Artificial Intelligence and Data Acceleration initiative, according to Sherman. Looking ahead, experts say the success of the JWCC depends on its ability to maintain flexibility and agility in the project.
“The cloud marketplace is actually relatively dynamic … the nature and capabilities of cloud are changing, and so you need to have the flexibility to identify what works for you at any given moment,” Soloway said. “And that’s what I think this, hopefully, or theoretically, should give them.”
Sherman already indicated Tuesday the department expects the other major cloud players not specifically named in the JWCC announcement—IBM, Oracle, and Google—might eventually be folded into the project. Sherman told reporters he would be reaching out directly to those cloud service providers. A Google spokesperson said in a statement the company looks forward to working with DOD as it continues to assess its IT needs, and an IBM official said in a statement the company is evaluating the new JWCC information. Oracle declined to comment.
Acton said ensuring flexibility in the JWCC project may need to go beyond simply pivoting to a multi-cloud structure. Acton wants to see a line-item in the contract that allows for innovation—particularly because the rapid pace of change in the cloud sector means it’s hard to predict which player or what technology is going to be needed years down the road
“I think what they need to do with JWCC is have some kind of clause in the contract, or some kind of ongoing innovation investigation to see what new cloud capabilities are well suited for DOD and have some kind of option to add them to the contract,” he said.
At the end of the day, JEDI may be one of the most prominent examples of a problem procurement experts have acknowledged for quite some time: that the technology sector is dynamic, and DOD and the federal government writ large have not yet adapted acquisition practices to be able to keep relative pace with industry.
“The idea that you can lock in a 10-year contract or a five-year contract with one vendor and be certain that you’re going to be getting the best solutions at the moment defies the logic of what we have going on in technology, which is changing moment to moment,” Soloway said.