WASHINGTON — “I don't think that we had any way to know how much pushback there would be.”
Chris Lynch is sitting in his office, wearing shorts and a hoodie, gesturing to the walls and popping mints into his mouth at a rapid rate. The founder of the Defense Digital Service is just weeks away from leaving the Pentagon after four years, a period during which he won respect from the various military branches and internal skeptics.
But in this case, Lynch isn’t talking about broad pushback to his office, located in the heart of the Pentagon but decorated and organized like nothing else in the DoD’s hierarchy. Nor is he talking about any resistance to his leaving, to be replaced by Brett Goldstein later this month.
He’s talking about one program, maybe the most infamous that his team has had to deal with: the Joint Enterprise Defense Initiative, or JEDI, a cloud computing contract that should form the basis of the department getting a workable cloud architecture up, but has been mired in lawsuits and PR assaults from almost the beginning.
Estimated to be worth $10 billion over its lifespan, JEDI went through a downselect that concluded only Amazon Web Services and Microsoft could meet the requirements. Prior to those selections, but predicting a limited field, Oracle launched a lawsuit and an organized pressure campaign through both the press and the Hill, which eventually led to an internal Pentagon ethics investigation and a pause on the program.
Two companies have been selected to move forward in the Department of Defense's Joint Enterprise Defense Infrastructure cloud contract consideration.
With that investigation now finished, the contract could be awarded as early as mid-July. Critics have said that the JEDI contract is effectively tailored for an Amazon Web Services award, as few other companies are large enough or have the necessary security authorizations to meet the Pentagon’s needs. Defense Department officials counter they are running an open and fair procurement focused on getting the best tech to support the warfighter.
During the process JEDI process, many companies and experts questioned why the contract had to be awarded to a single provider, rather than a group of providers that could collectively meet the requirements. But Pentagon officials have said that a single provider would best meet the needs of the war fighter.
The scope of the pushback from industry caught the Pentagon team, and DDS, by surprise. Lynch admitted he was considering leaving the department earlier, but stayed in order to make sure JEDI made it through 2018, when the contract seemed most vulnerable — or as he put it, “it has been very clear that the chances of success could be degraded if there was a disturbance in the force, so we just sort of decided to double down.”
That attitude writ large is how DDS got through the JEDI wars of the last year, one that might summarize the office’s attitude toward large-scale issues generally.
“I’ve said, you know, how lucky and fortunate are we that we get to work on things that matter, that people care so much about even if they are,” Lynch said. “We had to just decide that we were going to be happy with it and embrace it and we were going to be okay with it. We decided that, and I think that I wouldn’t have changed a single thing. It continues to go — it’s not done, but we’re on the right path.”
The official Department of Defense cloud strategy will pursue a single general purpose cloud surrounded by multiple other cloud environments designed for more specific uses.
That path goes back to a thought exercise in the early years of DDS. As Lynch describes it, his office didn’t set out to get the Pentagon on the cloud. Rather, the goal was to figure out how to provide “a capability that today doesn't exit in the department, that has severe consequences for how we do our mission.” They eventually settled on cloud computing as a method, and JEDI as the plan of attack.
The goal was to “go to that future and then work your way back,” he said. Conceptually, that meant thinking about where the Pentagon’s technologies are lacking, envisioning what a realistic tool to fill a gap might be, and then going step by step backwards to figure out how to build that out.
In the JEDI case, it started with the question of whether the Pentagon could develop something “where somebody can write things and it will run on a classified system and their development environment can go up to a secret system or a top secret system,” Lynch said. “That whole thing could be flawless and they can deploy anywhere around the world at any period of time. That is a strong vision. We have nothing like that today.
“That has nothing to do with commercial cloud, it has nothing to do with JEDI, right? But we don't have it today, that's actually a fantasy. So what if we had that?”
Still, he’s more philosophical than disgruntled by the situation.
“Yeah it’s been, I guess, surprising, but I suppose at the end of the day we always say this: nothing comes for free,” Lynch said. “You can’t make a substantial change at the Department of Defense without pushing really hard on some things that need to be pushed. I suppose that that’s the most amazing part about it, is that it matters so much that it has actually gained the attention of so many people.”
In a letter to the DDS team last week, Lynch wrote about the importance of the upcoming contract.
“Incredible things are ahead of you. You have tackled some of the DOD’s biggest challenges and have never had more support across the department,” he wrote. "JEDI is coming (the nerds have won my friends). And yes, we will continue to drive JEDI forward as we have since the start to better protect the young men and women who keep us safe.
Jessie Bur with Federal Times contributed to this story.