Why Industry Should Not Worry about the DOD's JEDI Cloud

The industry outcry has been nearly constant since last year’s announcement that the Department of Defense intends to award a contract for a Joint Enterprise Defense Infrastructure (JEDI). Commonly billed as the “one cloud to rule them all,” combining Lord of the Rings and Star Wars metaphors, JEDI has been the subject of significant lobbying by companies on Capitol Hill seeking either to stop the procurement altogether, ensure that it will be a multiple award contract, or, if the contract is to be a single award, make sure it doesn’t go to Amazon Web Services, which is widely considered the leading candidate thanks to its experience building a similar cloud for the Intelligence Community.

As an industry analyst, I’ve chosen not to weigh in on the subject because a number of news outlets have already asked my opinion and I’ve thought that sufficient. Now, however, I believe the time has come to explain why – in as much as it might matter to anyone out there what I think – industry fears about JEDI will likely not be realized and why a single contractor for JEDI, if indeed that is what the DOD intends to do, may not be the “Amazon set-aside” some may think.

The available evidence points to the conclusion that the JEDI cloud is intended primarily to support the warfighting efforts of DOD’s subordinate military departments. In testimony he gave before the House Armed Services Committee on the 26th of April 2018, Secretary of Defense James Mattis explained that, “DOD must remain on the cutting edge of advanced computing capabilities to support warfighting and lethality (my emphasis). Our cloud initiative simplifies the ability to provide enterprise-wide access to information and improves security to safeguard critical information.” Mattis has commented elsewhere as well that the JEDI program is central to the new 2018 National Defense Strategy. Furthermore, the JEDI draft Statement of Objectives makes it abundantly clear that the cloud services provided will synthesize every type of intelligence information out there (i.e., multi-INT) to provide decision-making support for commanders and warfighters on the tactical edge globally.

In short, from all appearances JEDI will be a tactically-oriented combat support system that fundamentally weaponizes cloud computing capabilities. As such, its use in a warfighting capacity sets parameters around the type of functions that JEDI will fulfill. These functions do not on the surface of things include financial systems, logistics, personnel and pay, health care, R&D, or any of the other myriad uses to which DOD puts its computers currently.

If you are a combatant commander in southwest Asia do you need data provided by a commercial cloud to tell you how research on artificial intelligence is progressing? The answer is probably no. You need an AI capability to sort through all of the video feeds coming in and synthesize it with other INT sources to help you make an informed operational decision.

The intent behind JEDI that’s been revealed so far actually reminds me of a cloud capability that the DOD already possesses called “Rainmaker.” Several years ago the Army, with the help of Booz Allen Hamilton, built an intelligence gathering cloud called Rainmaker to compile, synthesize, and provide analysis of intelligence data gathered through the DOD’s Distributed Common Ground System (DCGS). Analysis results are then pushed to combatant commanders engaged in operations around the world.

It sure sounds to me like JEDI will be Rainmaker’s big brother. If that is the case then even a multi-billion dollar single contract award will not stop billions of additional dollars being awarded for work on other cloud efforts across the DOD. Proof of this can perhaps be found in the May 2018 memo from the acting DOD CIO ordering Defense Agencies to begin moving systems to the milCloud 2.0 architecture that General Dynamics IT’s new CSRA subsidiary has built for the Defense Information Services Agency. Similarly, the Navy is proceeding with its own enterprise cloud architecture procurement, the Army continues to invest in its contractor owned and operated cloud at Redstone Arsenal, and work on the Air Force’s Installation Processing Node initiative is still going on. In other words, work is not stopping on cloud infrastructure efforts already underway at the DOD and more are being announced every day.

Add to all of this a telling comment made by Robert Daigle, Director of Cost Analysis and Program Evaluation at the Pentagon, a few weeks back. Daigle said, “Folks are trying to present this as the Department of Defense moving to a single award for all cloud computing. This contract, even if we max out the annual spend, represents about 16 percent of our hosting costs currently, which represent about 12 or 13 percent of our overall business system IT costs. So really the amount that we’re talking about here is not the end of history; it’s one cloud contract.”

All of this leads to the conclusion that JEDI will be used for purposes related primarily to combat support. Such a use will inherently limit the JEDI cloud’s functions and the type of data it is intended to store. Non-combat system/intelligence related data will still need to be housed in other commercial clouds to reduce costs and acquire for DOD the best new commercial cyber security capabilities that are available. It doesn’t appear that cloud computing competition at DOD will be at risk.