Earlier this month, the Department of Defense (DoD) submitted a report to Congress detailing the department’s justification for making its $10 billion Joint Enterprise Defense Initiative (JEDI) cloud contract a single award. Specifically, the report, which was mandated by Congress part of the Fiscal Year (FY) 2018 National Defense Authorization Act (NDAA), is intended to provide additional details explaining DoD’s acquisition strategy for the JEDI vehicle, particularly, why the department believes that it is in the government’s best interest to pursue a single-award approach for a contract lasting up to 10-years for potentially billions of dollars in IT infrastructure services.
Although the Coalition commends DoD for its transparency in making its decision/rationale publicly available, the report raises additional questions and concerns related to the department’s single-award strategy. This week’s blog focuses on key questions arising from DoD’s report that stakeholders should understand and request the department answer.
When considering a multiple- vs single-award approach, DoD said that it elected to forgo a multiple-award approach because it would not provide the department with the acquisition speed it desires. Specifically, the report states:
Under current acquisition law, if the Department pursued multiple-award contracts for the JEDI Cloud, each individual task order would be competed, thus being paced by DoD acquisition processes. That pace could prevent DoD from rapidly delivering new capabilities and improved effectiveness to the warfighter that enterprise-level cloud computing can enable.
This rationale, however, at best, is unconvincing, as it fails to recognize that tasks orders are designed to be conducted in streamlined manner so that the government can yield the benefits of competition without unnecessarily delaying the delivery of end solutions. In addition, pursuant to the fair opportunity requirements found at FAR 16.505(b)(2), DoD currently has the authority to streamline task orders by establishing specific requirements on a task order that will be competed. Moreover, these decisions are not subject to protest for task orders valued at less than $25 million, and DoD still reserves the right to utilize their Competition in Contracting Act (CICA) authorities to pursue separate contracts, if need be.
In addition, on page 4 of its report, DoD introduces the concept of a “data lake,” to further rationalize a single-award approach for its Cloud strategy. Specifically, DoD states:
Requiring multiple vendors to provide cloud capabilities to the global tactical edge would require investment from each vendor to scale up their capabilities, adding expense without commensurate increase in capabilities. … Maintaining inconsistent and non-standardized infrastructures and platform environments across classification levels complicates development and distribution of software applications, potentially adding delays and costs. Use of multiple clouds would inhibit pooling data in a single cloud (i.e., a “data lake”), limiting the effectiveness of machine learning.
DoD’s use of the data lake as a metaphor is significant in that it makes DoD’s reasoning appear to be plausible, but when examined further, it quickly raises questions regarding the approach. Specifically, DoD’s argument relies upon the assumption that commercial vendors do not have the capabilities they are describing. These vendors, however, supply the cloud solutions for the private sector, and these solutions routinely operate multiple cloud environments. Further, DoD does not explain how it will address the risks associated with a single location of data (or, to use DoD’s metaphor, what happens if the data lake runs dry or becomes contaminated). Moreover, the report does not provide substantive detail as to how a single point of failure would better facilitate resiliency, as called for by the Joint Requirements Oversight Council (JROC), relative to a scenario with multiple such points.
The Coalition also is concerned with DoD’s reasoning in the report related to the JEDI Cloud Solution’s two-year initial base period of performance. On page 11 of the report, DoD states that it will only exercise options on the solutions beyond the two-year initial base if the solution “is the most advantageous method” for the government in meeting its requirements. This rationale, however, does not align with DoD’s own statements made just two pages prior in the report. Specifically, on page 9, DoD quotes a Gartner study providing, “Much like the transition from mainframes to PCs, the transformation to cloud computing and hybrid IT architectures should be seen as a multiyear evolutionary process.” [Emphasis Added] Consequently, DoD’s reasoning that it simply can walk-away from the JEDI solution in two-years, at best, is unrealistic, and it is inconsistent with the department’s position.
Finally, the Coalition would like to touch upon the Gartner study quoted in DoD’s report. In particular, because of the absence of a full citation, the Coalition is unable to locate the report that DoD used to support of a single-vendor approach for Cloud solutions. We, however, were able to find an October 2016 report from the firm that notes, “Gartner expects multiprovider IaaS/PaaS strategies to become the de facto standard.” In this regard, it was interesting to note that, just this week, in a hearing before the House Oversight and Government Reform Subcommittees on Information Technology and Government Operations, DoD’s new CIO commented that, “in a cloud world, there is no such thing as one solution is going to solve for all.”
Given the immense value and performance period of the contract, as well as the critical needs of DoD, the Coalition believes that the department’s decision needs to be understood by all stakeholders and compliant with fundamental laws in the government market space. At this point, the report appears to raise more questions than it has answered.