General Dynamics Information Technology, Inc. v. United States

Opinion

In the United States Court of Federal Claims

GENERAL DYNAMICS INFORMATION TECHNOLOGY, INC.,

Plaintiff,

v. No. 26-cv-292

THE UNITED STATES, Filed Under Seal: August 10, 2026

Defendant,

Publication: August 22, 2026 and

PERATON, INC., Intervenor-Defendant.

Noah Bleicher of Jenner & Block, LLP, Washington, D.C., appeared for Plaintiffs. With him on the briefs were Moshe B. Broder and Jennifer Eve Retener of Jenner & Block, LLP, D.C.

Sheryl Floyd of the United States Department of Justice, Senior Trial Counsel, Civil Division, Department of Justice, Washington, D.C. appeared for Defendant. With her on the briefs was David Lank, Office of the General Counsel, United States Department of Health and Human Services.

Jeffrey M. Lowry of Vedder Price P.C., Washington, D.C., appeared for Intervenor-Defendant. With him on the briefs were Kevin Paul Connolly, Kelly E. Buroker, and Michael Ols of Vedder Price P.C., Washington, D.C.

2

MEMORANDUM AND ORDER

This post-award bid protest challenges Defendant Department of Health and Human Services (HHS) Centers for Medicare and Medicaid Services’ (CMS’s or Agency’s) issuance of a task order to Peraton, Inc. (Peraton) for Durable Medical Equipment (DME) Claims Processing support. ECF No. 34 (Plaintiff’s MJAR) at 9. In the course of the procurement, both Peraton and Plaintiff General Dynamics Information Technology, Inc. (Plaintiff or GDIT) received equally high ratings across the non-price factors CMS applied; however, Peraton’s offer was $16 million less than GDIT’s offer. GDIT argues that a substantial factor in Peraton’s win was that, rather than hosting the DME environment at a data center Peraton owned, containing a physical mainframe it operated itself, Peraton proposed contracting out this mainframe service or alternatively using Defendant’s own data center. The Solicitation, on its face, did not permit the use of Defendant’s data center.

GDIT challenges the Agency’s award to Peraton on multiple grounds, including that the Solicitation bars the use of contractor-run data centers like Peraton’s; that the Solicitation did not reflect CMS’s needs regarding the data center; and that the evaluation was unfair and irrational. Part of GDIT’s claim of irrational evaluation rests on portions of the Administrative Record indicating that CMS used incorrect, inflated numbers in its analysis of part of Peraton’s proposal. GDIT also moves to complete or supplement the Administrative Record, based on its claims that there are purportedly documents reflecting that the Agency planned to, and did, transition the DME environment to Defendant’s data center in contravention of the Solicitation. For the reasons stated below, the Court DENIES Plaintiff’s Motion to Complete and Supplement the Administrative Record (ECF No. 35), GRANTS IN PART and DENIES IN PART Defendant’s Motion for Judgment on the Administrative Record (ECF No. 43), GRANTS IN PART and DENIES IN

3

PART Peraton’s Motion for Judgment on the Administrative Record (ECF No. 41), GRANTS IN PART and DENIES IN PART Plaintiff’s Motion for Judgment on the Administrative Record (ECF No. 34), and REMANDS this action to CMS to reevaluate Peraton’s proposed Level of Effort as well as the best value determination to the degree it is impacted by such reevaluation.

BACKGROUND

I. Factual Background This bid protest challenges CMS’s issuance of a task order to Peraton for Durable Medical Equipment (DME) Claims Processing support under Solicitation No. 75FCMC26Q0007 (Solicitation). Pl. MJAR at 9; AR 490. “Durable Medical Equipment, Prosthetics, Orthotics, and Supplies (‘DMEPOS’) are medical items such as wheelchairs, hospital beds, and oxygen equipment covered under Medicare Part B.” Pl. MJAR at 11. CMS relies on the DME fee-for- service (FFS) claims processing environment to process Medicare claims related to these items; the DME FFS is one of CMS’s four claims processing “Shared Systems.” Id.; see also AR 784– 85, 804–05. GDIT is the incumbent contractor supporting the DME FFS claims processing environment. AR 50. The other three Shared Systems (not at issue here) are the Fiscal Intermediary Shared System (FISS) for institutional Part A and B services, Multi Carrier System (MCS) for physician and other practitioner services, and Common Working File (CWF) for sharing data with contractors and verifying beneficiary eligibility. AR 785. Peraton is the incumbent contractor for operation and maintenance of the FISS and MCS Shared Systems. AR 1195.

CMS’s mission, within the context of the contract at issue here, includes the following objectives:

4

1. Ensure a stable Medicare [FFS] claims processing environment, 2. Modernize CMS’s systems and databases and reduce system maintenance costs while extending the life of the system, 3. Move toward standardizing all [Shared System Maintainers (SSMs)], 4. Improve system quality and performance, 5. Incorporate legislative and CMS mandates, and, 6. Implement changes and enhancements.

AR 784.

The Solicitation specified that the procurement would be held in accordance with FAR 8.405. AR 695. It also included a Statement of Objectives (“SOO”) with which the contractors were to comply. AR 786. According to the SOO, the “primary and over-arching objectives CMS has for this contract are to: . . . Measurably increase, over the period of performance, the use of modern technology, including, but not limited to Cloud-based solutions, use of modern tools such as, Jira and Confluence [and] Contribute to CMS’s efforts to modernize . . . its fee-for-service ecosystem . . .” AR 788. Under the SOO, the contractor “shall provide a system environment that is capable of supporting the development, maintenance, testing and all associated activity of the DME system.” AR 786. “The DME claims processing shared system runs in an International Business Machines (IBM) operating environment.” AR 788. The SOO required the contractors to coordinate their work with the Hybrid Cloud Data Center (HCDC), “the data center for CMS data processing projects involving Medicare, Medicaid,” and additional associated programs for FFS “data collection, storage, statistics, and data integrity.” AR 803. In the appendix to the SOO, when discussing the HCDC, the Solicitation stated that “DME SSM shall support development and maintenance of the DME system at their data center.” AR 803 (emphasis in original).

The Agency assessed vendors based on price and three non-price factors: (i) Case Study Corporate Experience, (ii) Oral Presentation, (iii) Performance Work Statement (PWS)/Quality Assurance Surveillance Plan (QASP). AR 695. Case Study Corporate Experience was more important than Oral Presentation, and Oral Presentation was, in turn, more important than

5

PWS/QASP. Id. Quotes were evaluated for “low,” “some,” or “high” confidence. See AR 2940, 2089. All non-price factors, when combined, were “significantly more important than cost or price,” and the Agency would evaluate the quotes for the best value to CMS. AR 695. The Agency required all questions regarding the Solicitation to be emailed to the Agency by September 15, 2025. AR 696.

CMS adopted a multi-phase approach for submitting quotes to minimize quote development and presentation costs for respondents with a lower chance of receiving an award. Id. The Solicitation stated that CMS would make an “advisory down-select” between each phase. Id. Any respondents advised not to continue were allowed to participate in the next phase if they elected to do so and advised CMS of their intent to participate. Id. Phase 1 began with respondents submitting proposals for factor 1 (Case Study Corporate Experience) and then proceeding to Phase 2 for evaluation of all other factors. Id.

Phase 1 was a case study where respondents described the context surrounding a past project in which the respondent “leveraged modern technology or practices to improve and modernize a claims processing system.” AR 698. Respondents described a recent experience, within the past three years, serving as a prime contractor or contractor teaming arrangement (CTA) teammate that contracted with federal, state, or local agencies or commercial contractors. AR 698–99. The Agency then assessed how well each respondent’s previous experience demonstrated performance or work similar to the work the Agency sought in this procurement. AR 699-700.

At Phase 2, respondents answered interview-style core questions at the oral presentation.

AR 700. They also responded to hypothetical Agency problems with hypothetical solutions. Id. The Agency assessed the respondents’ performance and answers based on their “capability and suitability” to perform the work. AR 702. Phase 2 involved the respondents proposing a

6

Performance Work Statement (PWS) and QASP. Id. The PWS detailed respondents’ understanding of CMS’s needs and provided a proposed solution. Id. Specifically, the PWS needed to detail: the task to be performed and deliverables to be provided, the people, tools, methods and management of the effort, and underlying assumptions. Id. It also needed to include an organizational chart or staffing plan, the contractor’s own estimated skill mix, 1 level of necessary effort, proposed hiring plans and timelines, and a list of assumptions, exclusions, exceptions, and clarifications. AR 702–03. The QASP illustrated how the proposed performance standard would be “measured, evaluated, and reported” and would be used by the Agency “in developing its own QASP.” AR 703. The Agency based its evaluation of the PWS/QASP element on how well it conformed to the objectives in the SOO, and whether the PWS demonstrated the respondent’s “understanding of the [Agency]’s needs, and the extent to which the people, processes, performance measures and tools [would] serve to address the needs and challenges associated with performance of the work.” Id.

For price, the final factor, vendors were required to include a basis of estimate (BOE), including the estimating method used to develop the proposed price, the types of labor costs, materials or performance information, and rationales for proposed quantity. AR 707–08. The price spreadsheet included line-items for direct and indirect costs, as well as the breakdown of costs by material, type, and quantity. Id. The Agency would evaluate a respondent’s price for fairness and reasonableness, and the Solicitation noted that any inconsistency between the price spreadsheet and the rest of the quote was grounds for removal from consideration. AR 709.

1 The respondents were permitted to propose a level of effort (LOE) but were “not bound to quote only what is estimated or assumed by the Government in this regard”; rather, they were to “quote their own estimated skill mix and level of effort which they believe[d] necessary to deliver the approach described in the PWS.” AR 702.

7

The bidders submitted many questions to the Agency regarding the Solicitation’s requirements. See AR 2939. Among them were several requests for clarification regarding the data center requirement. For instance, one potential bidder asked: “The current Dev and Test Environment infrastructure may not be the most advantageous to CMS. Will the [Agency] consider cost savings approaches to the Dev/Test requirements by consolidation to the existing CMS CACHE 2 infrastructure during the period of performance of this contract?” AR 771 (DME Consolidated Questions and Answers or Q&A) at Question 7. The bidders lodged several other requests surrounding the data center requirement with the Agency. See, e.g., id. at Questions 10, 12–16, 33, 34. The Agency emphasized that it had not moved the data environment to the CACHE and would not consider it at that time. Id. at Question 12.

For instance, one respondent asked specifically whether the Appendix A instruction should “be interpreted to mean that the awarded SSM under this contract will provide a replacement data center, which the DME systems will migrate to,” because the inquiring respondent found the word “‘their’ in the previous Q&A a bit confusing.” Id. at Question 32. The Agency responded that “their” data center meant “the awardee[’]s” data center. Id. Similarly, one respondent inquired whether the Agency would “consider transitioning the VMS development environment to an AWS cloud environment in order to avoid vendor lock in and enable a smoother modernization effort?” to which the Agency responded, “No, not at this time.” Id. at Question 17.

Plaintiff and Peraton timely submitted factor 1 quotations for Phase 1 by the Solicitation’s deadline of September 22, 2025. AR 2938. Including Plaintiff and Peraton, nine companies

2 The CACHE is a “Continuously Availably CMS Hosting Environment” data center. See Peraton MJAR at 15–16.

8

submitted timely quotes for Phase 1. AR 2938–39. Subsequently, on November 24, 2025, Plaintiff and Peraton timely made their Phase 2 submissions. AR 2940.

Of special note, in its Phase 2 submission, Peraton included a PWS/QASP which stated:

While Peraton’s proposal includes a hosting solution, pending CMS approval, Team Peraton will collaborate with CMS to explore potential cost savings that can be achieved by transitioning the DME Dev/Test hosting into the CMS CACHE data center, building on prior successful FFS Dev/Test transitions to CACHE. Operating the DME Dev/Test environments within CMS’ CACHE data center provides opportunities to leverage CMS CACHE enterprise services and resources, optimize the use of software licenses and reap the benefits of economies of scale.

AR 1748.

The hosting solution involved using a contractor, , which would provide the physical mainframe infrastructure to host the DME on a monthly cost basis. See AR 1747, 1878.

[continued on next page]

9

The evaluation results across Phases 1 and 2 were as follows:

Factor 1 (Phase Factor 2 (Phase Factor 3 (Phase Total Quoted Quoter 1) (Case Study– 2) (Oral Two) Price 3 Corporate Presentations) (PWS/QASP)

Experience)

High High High $71,969,410.61 High High High $56,935,362.48 N/A N/A N/A

N/A N/A N/A

N/A N/A N/A

N/A N/A N/A

N/A N/A N/A N/A N/A N/A

AR 2943.

The Contracting Officer (CO) determined that the prices of all three contractors that proceeded into Phase 2 (Peraton, GDIT, and ) were reasonable. AR 2956-57. Peraton’s total quoted price was $56,935,362.48—significantly less than GDIT’s $71,969,410.61 price. AR 2945.

On February 6, 2026, CMS sent Peraton an email notifying it of its selection for the award and informing it that the performance period was expected to begin on Monday, February 9, 2026.

3 Excludes the six-month extension under FAR 52.217-8.

10

AR 3235. On February 9, 2026, the Agency officially awarded Peraton the contract. AR 3227. That same day, CMS informed Plaintiff of Peraton’s selection for the award. AR 3215. On February 11, 2026, Peraton requested a brief explanation “for the basis of award,” and “that the brief explanation also be held with Peraton telephonically.” AR 3252.

The Agency provided GDIT and Peraton a written explanation on February 17, 2026. AR 3246–51; AR 3253–57. The brief explanation included both Peraton’s and Plaintiff’s ratings: each received “High Confidence” ratings for all three non-price factors. AR 3247. The explanation specified that Plaintiff had received multiple confidence increasers and no confidence decreasing factors across all technical factors. AR 3247–51. However, as the brief explanation detailed, Peraton offered a price $16,493,001.53 lower than Plaintiff’s offer. See AR 3247. The brief explanation further stated that, after conducting a tradeoff analysis, the Agency determined that Plaintiff’s technical approach did not offer enough value to warrant the price premium. Id. Additionally, the brief explanation provided to Peraton repeated the evaluations of its scores on the three non-price factors and did not mention GDIT. AR 3254–57.

In a declaration attached to its original Complaint 4 but not to its Amended Complaint, GDIT alleges that on February 18, 2026, , program Senior Manager for GDIT’s Durable Medical Equipment (DME) contract, received a call from DME’s Contracting Officer Representative, . ECF No. 1-2 ( Declaration) at 173–74. According to the Declaration, “informed [ ] that CMS has decided to transition from the DME Data Center hosted by GDIT to the Continuously Available CMS Hosting Environment (CACHE) Data Center.” Id. at 173.

4 As noted, Plaintiff did not attach the Declaration to its Amended Complaint. See ECF No. 33-1 (Amended Complaint or Am. Compl.).

11

II. Procedural History On February 20, 2026, Plaintiff filed its original Complaint. ECF No. 1 (Complaint).

Plaintiff also moved for a Protective Order, which the Court granted on February 23, 2026. ECF No. 4; Minute Order dated Feb. 23, 2026, ECF No. 8. Subsequently, Peraton filed an unopposed Motion to Intervene, which the Court granted on February 25, 2026. ECF Nos. 12, 14.

The Court held an Initial Status Conference on March 9, 2026, and entered a Scheduling Order that same day. See Minute Entry dated Mar. 9, 2026; ECF No. 23 (Scheduling Order). Consistent with the Court’s Scheduling Order, on March 24, 2026, Defendant filed a Notice of File-Sharing of the Administrative Record. See Scheduling Order; ECF No. 24 (Notice of File- Sharing). On March 25, 2026, Defendant filed an Unopposed Motion for Leave to File Corrected Notice of File-Sharing of the Administrative Record seeking to add the Contracting Officer Certification and Administrative Record Index to the Administrative Record, as those documents “were inadvertently not attached to the Notice of File-Sharing official[ly] filed with the Court.” ECF No. 25 (First Motion to Complete) at 1-2. The Court granted the First Motion to Complete on March 27, 2026, and on March 31, 2026, Defendant filed a Corrected Notice of File-Sharing of the Administrative Record. Minute Order dated Mar. 27, 2026; ECF No. 26.

On April 1, 2026, Defendant inadvertently filed a duplicate of its First Motion to Complete titled “Defendant’s Unopposed Motion for Leave to File Corrected Notice of File-Sharing of the Administrative Record and of Filing the Contracting Officer Certification of Administrative Record and Administrative Record Index.” See ECF Nos. 27 (Duplicate Motion), 29 (April 6, 2026, Notice) (explaining the inadvertent filing). The Duplicate Motion contained a proposed attachment, Administrative Record (AR) 2913.1. ECF No. 27-3. In Defendant’s April 6, 2026, Notice, it advised the Court that on April 6, 2026, Plaintiff stated via email to Defendant that it did

12

not oppose Defendant’s addition of AR 2913.1 but “would oppose [] [Defendant’s] motion to amend and correct the AR because [] [Defendant] has declined to add any additional documents to the AR.” April 6, 2026, Notice at 3 n.2. Defendant’s April 6, 2026, Notice also requested that the Court grant it leave to file a corrected Motion to Amend and Motion to Complete the Administrative Record. Id. at 4. The Court accordingly struck the Duplicate Motion from the record and granted Defendant’s motion to refile. ECF No. 30 at 1–2. Defendant then filed a corrected Unopposed Motion to Amend and Motion to Complete the Administrative Record on April 10, 2026. ECF No. 31 (Second Corrected Motion to Amend and Motion to Complete the Administrative Record).

On April 14, 2026, Plaintiff filed its Amended Complaint. Amended Complaint. That same day, Plaintiff filed two motions: its Motion for Judgment on the Administrative Record, ECF No. 34 (Pl. MJAR), and its Motion to Complete and Supplement the Administrative Record, ECF No. 35 (Motion to Complete and Supplement or Mot. to Suppl.). The Motion to Complete and Supplement sought completion of the AR with certain materials and, alternatively, sought supplementation of the AR with certain materials related to actions CMS may have considered at or after the contract award. See generally Mot. to Suppl.

On May 5, 2026, Defendant and Intervenor-Defendant each filed respective Cross-MJARs and Responses to Plaintiff’s MJAR, as well as their Oppositions to Plaintiff’s Motion to Complete or Supplement. ECF Nos. 41 (Peraton MJAR), 43 (Def. MJAR), 42 (Peraton Opp. to Suppl.), 44 (Def. Opp. to Suppl.). On May 19, 2026, Plaintiff filed its Reply in support of Plaintiff’s MJAR and Response to Peraton and Defendant’s MJARs. ECF No. 45 (Pl. Reply). That same day, Plaintiff filed a Reply in Support of its Motion to Complete and Supplement the Administrative Record. ECF No. 46 (Reply ISO Suppl.). Defendant and Intervenor-Defendant filed their

13

respective Replies in support of their Cross-MJARs on June 3, 2026. ECF Nos. 47 (Peraton Reply), 48 (Def. Reply). On June 16, 2026, the Court held oral argument on the pending motions. See Minute Entry dated June 16, 2026.

DISCUSSION

I. Motions for Judgment on the Administrative Record Plaintiff asserts five arguments in its MJAR: (i) Defendant failed to recognize that Peraton’s proposal involved a radically low amount of labor hours; (ii) Peraton proposed a non- compliant data center solution that rendered Peraton ineligible for the award; (iii) the allegedly sudden post-award move toward using Defendant’s own data center (the “Continuously Availably CMS Hosting Environment” or CACHE) indicates that the Solicitation did not accurately describe the Agency’s needs; (iv) Defendant purportedly treated Peraton and Plaintiff differently by giving Peraton alone high marks for its QASP; (v) the best value evaluation was unreasonable. Plaintiff’s first argument is its most successful: the Agency erred in certain recitations of Peraton’s proposed labor hours, and it is unclear from the record whether the Agency considered the accurate amount of labor hours while assessing Peraton’s proposed Level of Effort (LOE) or in making a best value determination. See infra Discussion I.B.2, I.F.2. The Court finds Plaintiff’s remaining arguments unpersuasive. Accordingly, this matter is remanded to the Agency to (i) reevaluate Peraton’s proposed LOE using accurate labor hour figures, and (ii) perform a new best value determination consistent with this ruling and with the reevaluated proposed Level of Effort.

A. Legal Standards The Tucker Act, 28 U.S.C. § 1491(b)(1), as amended by the Administrative Dispute Resolution Act of 1996, affords this Court jurisdiction over bid protests. The Court analyzes the procurement under the Administrative Procedure Act’s (APA) standards to determine whether the

14

Agency “acted without rational basis or contrary to law when evaluating the bids and awarding the contract.” Bannum, Inc. v. United States, 404 F.3d 1346, 1351 (Fed. Cir. 2005); 28 U.S.C. § 1491(b)(4); see Oak Grove Techs., LLC v. United States, 116 F.4th 1364, 1374 (Fed. Cir. 2024). The APA requires a reviewing court to determine whether an agency’s action was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A); see Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 284 (1974). Thus, to prevail in a post-award bid protest, a plaintiff must demonstrate that “(1) ‘the procurement official’s decision lacked a rational basis’ or (2) ‘the procurement procedure involved a violation of regulation or procedure.’” DynCorp Int’l, LLC v. United States, 10 F.4th 1300, 1308 (Fed. Cir. 2021) (quoting WellPoint Mil. Care Corp. v. United States, 953 F.3d 1373, 1377 (Fed. Cir. 2020)). “[W]hen procedural violations committed by the agency are egregiously removed from fairness, this constitutes an abuse of the agency’s administrative discretion.” Progressive Indus., Inc. v. United States, 129 Fed. Cl. 457, 477 (2016) (quoting Doty v. United States, 53 F.3d 1244, 1251 (Fed. Cir. 1995)). “[U]neven treatment” of offerors may constitute a violation of the “standard of equality and fair-play that is a necessary underpinning of the federal government’s procurement process” and may “amount[] to an abuse of the agency’s discretion.” PGBA, LLC v. United States, 60 Fed. Cl. 196, 207 (2004) (citing Doty, 53 F.3d at 1251).

“Although the inquiry under the APA ‘is to be searching and careful, . . . [t]he court is not empowered to substitute its judgment for that of the agency.’” Insight Pub. Sector, Inc. v. United States, 161 Fed. Cl. 760, 786 (2022) (quoting Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 416–20 (1971)). As the Federal Circuit explained, “the disappointed bidder bears a heavy burden of showing that the award decision had no rational basis.” Centech Grp., Inc. v. United States, 554 F.3d 1029, 1037 (Fed. Cir. 2009) (quoting Impresa Construzioni Geom.

15

Domenico Garufi v. United States, 238 F.3d 1324, 1333 (Fed. Cir. 2001); see also Impresa, 238 F.3d at 1333 (noting a similarly high burden for claims of a violation of regulation or procedure, which must involve “a clear . . . violation of applicable statutes or regulations” (quoting Kentron Hawaii, Ltd. v. Warner, 480 F.2d 1166, 1169 (D.C. Cir. 1973)). Consistent with this high burden, agency decisions are “entitled to a presumption of regularity.” Impresa, 238 F.3d at 1338 (citing Bowen v. Am. Hosp. Ass’n, 476 U.S. 610, 626–27 (1986)).

Agencies possess “substantial discretion” to make decisions involving “the minutiae of the procurement process in such matters as technical ratings” or best value determinations. E.W. Bliss Co. v. United States, 77 F.3d 445, 449 (Fed. Cir. 1996) (explaining that “a court will not second guess” certain matters, such as “technical ratings . . . , which involve discretionary determinations of procurement officials”); Galen Med. Assocs., Inc. v. United States, 369 F.3d 1324, 1330 (Fed. Cir. 2004) (“[A]s the contract was to be awarded based on ‘best value,’ the contracting officer had even greater discretion than if the contract were to have been awarded on the basis of cost alone.” (quoting E.W. Bliss Co., 77 F.3d at 449)). Indeed, this Court’s “role in reviewing procurement decisions . . . is not to evaluate the offerors’ proposals anew or to substitute [its] judgment for that of the agency.” Harmonia Holdings Grp., LLC v. United States, 999 F.3d 1397, 1408 (Fed. Cir. 2021); see also Honeywell, Inc. v. United States, 870 F.2d 644, 648 (Fed. Cir. 1989) (“If the court finds a reasonable basis for the agency’s action, the court should stay its hand even though it might, as an original proposition, have reached a different conclusion as to the proper administration and application of the procurement regulations.” (quoting M. Steinthal & Co. v. Seamans, 455 F.2d 1289, 1301 (D.C. Cir. 1971))).

Even if the Court finds that the agency acted contrary to law or without rational basis, it must then examine whether the plaintiff has been prejudiced by the error. Sys. Stud. & Simulation,

16

Inc. v. United States, 22 F.4th 994, 998 (Fed. Cir. 2021). “We first ask ‘whether the agency’s actions were arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law’; if so, we ask whether the error was ‘prejudicial.’” Id. at 997. (cleaned up) (quoting Off. Design Grp. v. United States, 951 F.3d 1366, 1371 (Fed. Cir. 2020)). A protestor establishes prejudice by showing “that there was a ‘substantial chance’ it would have received the contract award but for” that error. Id. (quoting Bannum, 404 F.3d at 1353).

In the Court of Federal Claims, bid protests are adjudicated under Rule 52.1(c), which provides an expedited trial on a “paper record, allowing fact-finding by the trial court.” Bannum, 404 F.3d at 1356; see Rule 52.1(c). Unlike at summary judgment, genuine disputes of material fact do not preclude a court from granting a motion for judgment on the administrative record. Bannum, 404 F.3d at 1357. Further, this Court is empowered to provide any relief, including declaratory or injunctive relief, that it deems proper. 28 U.S.C. § 1491(b)(2); Oak Grove, 116 F.4th at 1375. To that end, if necessary, this Court may remand the case back to a governmental agency for further factual findings. See Rule 52.2.

B. Whether CMS Irrationally Evaluated Peraton’s Proposed Level of Effort Plaintiff argues that several errors in CMS’s evaluation of Peraton’s proposed Level of Effort (LOE) caused CMS’s award decision to be arbitrary and capricious. The Court agrees: both CMS’s Award Memorandum and Technical Evaluation Panel (TEP) report reveal significant, material discrepancies in the amount of total labor hours that CMS examined in its evaluation of Peraton’s proposed LOE, prejudicing GDIT. Accordingly, as noted, this action must be remanded to the Agency for reevaluation using accurate figures for Peraton’s proposed labor hours estimate.

1. Parties’ Arguments Plaintiff argues that CMS irrationally evaluated Peraton’s proposed LOE because Peraton’s proposal, Defendant’s TEP evaluation, and Defendant’s Award Memorandum contain

17

inconsistencies, including an incorrect (and significantly higher) labor hours estimate for Peraton’s proposal in CMS’s evaluation record.

First, Plaintiff argues that CMS irrationally evaluated Peraton’s proposed LOE in its PWS by missing inconsistencies in its proposal. Pl. MJAR at 28 n.15. Peraton’s proposal indicates different amounts of total labor hours in different places: “Peraton’s price spreadsheet reflected 466,066 5 (AR 1927 at ‘Labor Roll Up’ tab (Column H)), but the price narrative tallied 466,701.75 hours (AR 1187), and the awardee’s PWS listed 250.1 [Full-Time Equivalents (FTEs)], which calculates to 466.686.6 hours (AR 1757).” 6 Id. at 28 n.15 (emphasis in original). Plaintiff asserts that these inconsistencies are unexplained in the record and should have caused disqualification or a lower rating for Peraton. Id. Defendant argues these are “relatively small inconsistencies.” and proposes Peraton’s disclosed rounding assumptions as a potential source of the variations. Def. MJAR at 32–33. Defendant contends these are minor errors and asserts CMS is permitted to “waive informalities and minor irregularities in offers received.” Id. at 34 (quoting FAR 52.212- 1).

Aside from these inconsistencies within Peraton’s proposal, Plaintiff also “maintain[s] that Peraton’s exceedingly low level of effort is a risk to successful performance.” Pl. MJAR at 29. Plaintiff highlights that using the “466,066” figure from Peraton’s price spreadsheet resulted in “about 12 fewer FTEs per year on the program than what GDIT proposed,” or “a more than 20% reduction in the contract workforce.” Id. at 28–29. GDIT claims that a rational evaluation of the LOE should have detracted from Peraton’s rating because of the potential risk posed by its “too

5 The total in the “Labor Roll Up” tab amounts to 466,065.8, which Plaintiff appears to have rounded up to 466,066. See AR 1927 at ‘Labor Roll Up’ tab (Column H)). 6 The contractors each assigned different numbers of hours to their FTEs. See AR 2956.

18

few labor hours.” Id. Defendant responds that there was only a “moderate difference” between GDIT’s and Peraton’s proposed price volume for the LOE. Def. MJAR at 32 (citing AR 2945). The difference between its own estimate and GDIT’s, Peraton argues, is entirely to be expected with different PWSs containing different bases of estimates. Peraton MJAR at 34. Furthermore, Peraton emphasizes that the in its labor mix would rationally explain how a lower amount of labor hours would not necessarily create more risk than GDIT’s proposal. Id. at 34 (citing AR 2118, 2120).

However, the central error, Plaintiff claims, is that CMS’s assessment of Peraton’s LOE is internally inconsistent and relies upon incorrect, inflated labor hours estimates. Pl. MJAR at 28– 30. Namely, portions of the TEP Report and Award Memorandum erratically invoke an estimate of Peraton’s total labor hours of about 547,541 hours—higher than the amount Peraton actually proposed, which was approximately 466,066. Id. at 28–30. Plaintiff argues that it was prejudiced by this error. Id. at 44.

Defendant, for its part, acknowledges the error in the TEP’s references to the 547,541 figure. Def. MJAR at 33 (emphasis omitted). However, Defendant argues that this error is inconsequential because it allegedly does not speak to any internal inconsistency within Peraton’s proposal and the TEP was thus accurate in confirming there was no internal inconsistency. Id. (citing AR 2121). Peraton argues that the Contracting Officer did not rely on the incorrect labor hours estimate, as evidenced by the Award Memorandum’s use of the accurate total for Peraton’s proposed labor hours in parts of its evaluation. Peraton MJAR at 35–36, see AR 2956 (chart containing the correct figures), 2964 (citing the inaccurate figure), 2964–65 (including a chart with FTEs corresponding to the correct LOE). Further, Peraton argues, the Award Memorandum contains a justification for its positive evaluation, using a more accurate figure, in a summary prior

19

to the recitation of mistaken figures. Id. at 37 (citing AR 2963 (“The TEP evaluated labor hours, considering the approach Peraton laid out in their submission, and noted the following: The proposed 466,701.80 hours are adequate and appropriate for the work to be performed in accordance with the quoter’s approach.”)).

The Court agrees with Plaintiff. Plaintiff has succeeded in establishing a material contradiction in the record tainting Defendant’s LOE assessment, rendering the award to Peraton arbitrary and capricious.

2. Conclusion

CMS’s reasoning for its evaluation of Peraton’s LOE is not discernable from the record, and accordingly the LOE labor evaluation must be remanded to the Agency for reevaluation.

As an initial matter, the Court disagrees with GDIT that the Contracting Officer should have determined that Peraton’s proposed 466,000 labor hours proposal was “exceedingly low” and “a risk to successful performance.” Pl. MJAR at 29. Plaintiff’s argument comes down to an observation that Peraton’s proposed LOE is around 20% lower than GDIT’s, without any evidence that this labor reduction is untenable. Id. Indeed, as the Award Memorandum notes, while Peraton quoted fewer hours overall, it quoted a , which increased its capability. Compare AR 2962 (“[GDIT’s] staffing composition of in the base period demonstrates a strong emphasis on . . . personnel who bring expertise and capabilities.”), with 2963 (“[Peraton’s] staffing composition of in the base period demonstrates a strong emphasis on personnel who bring deep expertise and proven capabilities.”). Without any further evidence of the LOE’s unsuitability for performance in the record, the Court may not make the determination for CMS whether Peraton’s proposed LOE is realistic for the work to be performed. After all, in “best value procurements like this one, the

20

[Agency] has substantial discretion in its technical evaluations.” Newimar S.A. v. United States, 160 Fed. Cl. 97, 130 (2022), aff’d, No. 2022-1949, 2023 WL 8534614 (Fed. Cir. Dec. 11, 2023) (citing E.W. Bliss Co., 77 F.3d at 449); DevTech Sys., Inc. v. United States, 176 Fed. Cl. 297, 328 (2025) (listing cases).

However, it is undisputed that the LOE evaluation contains mistakes; namely, while Peraton proposed around 466,066 hours, the TEP evaluation and Award Memorandum sometimes cite a total labor hours estimate inexplicably higher than what Peraton actually proposed. OA Tr. 93:8–10 (Defendant’s Counsel: “So, yes, in fact, there were on more than one occasion did CMS have the wrong number, I don’t know why, if you ask, but it was an error.”), 110: 1–4 (The Court: “You agree that there are mistakes?” Peraton’s Counsel: “Correct, Your Honor . . . We have gone back through to try to figure out where this came from.”); see OA Tr. 90:1–13, 100:17–21. In the TEP Report, for instance, the TEP’s discussion of Peraton’s LOE cites Peraton’s “547,541 hours”– a figure which has no basis in Peraton’s actual proposed labor hours estimate. AR 2121. In fact, the TEP’s evaluation of Peraton does not use any variation of the lower total hours estimates actually in Peraton’s proposal—not 466,066 (AR 1927 at ‘Labor Roll Up’ tab (Column H)), not 466,701.75 hours (AR 1187), and not 466.686.6 hours (AR 1757). AR 2120–22.

More contradictions follow in the Award Memorandum. In the Award Memorandum, the CO states that “[t]he TEP evaluated labor hours, considering the approach Peraton laid out in their submission, and noted the following: The proposed 466,701.80 hours are adequate and appropriate.” AR 2963. This is inaccurate, as the TEP Report never uses the 466,701.80 figure. See AR 2120–22. Peraton argues that this adoption of the TEP labor evaluation provides sufficient justification of the labor evaluation in the Award Memorandum. Pl. MJAR at 37 (citing AR 2963). Indeed, the CO relies explicitly on the TEP Report in her findings. See AR 2956 (“The Contracting

21

Officer used the technical knowledge and expertise of the Technical Evaluation Panel (TEP) to determine whether the respondent’s quoted labor hours were reasonable.”). However, this adoption of the TEP’s rationalization will not suffice: not only does the CO misstate the TEP Report, but the TEP Report only ever cites the faulty 547,541 figure instead of the lower, approximately 466,000, figure. See AR 2121. Thus, insofar as the Award Memorandum adopted the TEP Report, it adopted an LOE evaluation based on an inflated number of labor hours for Peraton; accordingly, that the Award Memorandum “relied on the TEP’s analysis” only undercuts the rationality of the award determination. AR 2946.

Furthermore, there are too many material, internal contradictions in the Award Memorandum itself to overlook the errors. The Award Memorandum oscillates between the accurate and inflated labor hour total. At the start, the Award Memorandum initially lists the correct labor hours for all contractors:

[continued on next page]

22

Hours FTEs

Period GDIT Peraton Period GDIT Peraton Base Period Base Period Option Period Option Period One One Option Period Option Period Two Two Option Period Option Period Three Three Option Period Option Period Four Four Total 584903.3 466701.8 Total 311.12 250.11

AR 2956.

However, the Award Memorandum then alternates between citing the correct hours amount (or its FTE equivalent) and citing the inflated figure (or its FTE equivalent) in both its narrative analysis and illustrative chart. AR 2963–66, 2964 (“The total Labor hours are . . . for the base and 547,540.5 total labor hours for the base and optional performance periods.”). As stated, the Award Memorandum inaccurately described the TEP Report: “[t]he TEP evaluated labor hours . . . and noted the following: The proposed 466,701.80 hours are adequate and appropriate.” AR 2963. It relies upon a TEP Report which used an inflated number of hours. AR 2121. Then, the Award Memorandum uses an inaccurate 547,540.5 figure in its analysis of Peraton’s labor hours estimate. AR 2964 (citing the inaccurate figure). The Award Memorandum then includes a statement that Peraton’s hours “increase[] in option period four with the most hours,”—again, incorrect, as Peraton proposed the fewest hours in option period 4. Pl. MJAR. at 30 (citing AR 2964, 1887). This means that the Contracting Officer “assumed that Peraton proposed nearly 90 FTEs in the final year of performance when the vendor really proposed half that number.” Id. As a result, Plaintiff argues, the evaluation of Peraton’s LOE misunderstands Peraton’s proposal and is inconsistent with the record. Id. at 30–31.

23

The Agency’s use of the incorrect 547,541 figure in the TEP Report and 547,540.50 figure in the Award Memorandum presents significant material issues for CMS. In reviewing the Award Memorandum, it is difficult to discern which figures exactly the CO evaluated or approved: the document contains both. AR 2963–66 (“The total Labor hours are . . . for the base and 547,540.5 total labor hours for the base and optional performance periods.”); see OA Tr. 90:14– 16 (Defendant’s Counsel: “So [the Agency] uses two numbers, you don’t know looking at them that they are—you know, one of them is correct and one of them isn’t correct.”). While the Court has no evidence indicating that the lower number was per se unsuitable for performance, see supra at 19, it does not appear to be an insignificant change: the difference was that of around 15% less labor than both what the TEP thought Peraton had proposed or GDIT’s proposal, which itself was a decrease from its incumbent hours amount. See Def. MJAR at 29; AR 2963 (citing the correct estimate), 2964 (citing the inflated estimate). A rational basis for a contracting officer’s decision necessarily requires an understanding of material aspects of a proposal. Ala. Aircraft Indus., 586 F.3d at 1375; Barbaricum LLC v. United States, 172 Fed. Cl. 186, 197 (2024) (“Courts evaluating agency action on technical subjects therefore look to whether the record contradicts the agency’s reasoning in ways that cannot be ascribed to the agency’s judgment or expertise.” (citing DZSP 21, 139 Fed. Cl. at 118 n.9; Allicent Tech., LLC v. United States, 166 Fed. Cl. 77, 115 (2023))). Here, the Contracting Officer’s evaluation was obviously and materially inconsistent with the record and therefore is unreasonable. See NMB Singapore Ltd. v. United States, 557 F.3d 1316, 1319 (Fed. Cir. 2009) (“[W]hile its explanations do not have to be perfect, the path of [the agency’s] decision must be reasonably discernable to a reviewing court.” (citing State Farm, 463 U.S. at 43).

24

a) Prejudice

The Court also finds that Plaintiff has demonstrated it was prejudiced by the error. See Steel Point, 180 Fed. Cl. at 470 (“A protestor cannot prevail if it does not establish prejudice.” (citing DynCorp Int’l, LLC v. United States, 10 F.4th 1300, 1308 (Fed. Cir. 2021)).

Regarding the smaller inconsistencies with the approximately 466,066 hours figure, both Plaintiff and Defendant agree that there were at least some internal inconsistencies in the labor hours figure in Peraton’s LOE proposal. See Pl. MJAR at 28 n.15; Def. MJAR at 32; see also supra at 17 (describing the variations among AR 1927 (totaling 466,065.8), AR 1187 (using 466,701.75), 1757 (listing FTE’s resulting in 466.686.6 hours).” However, these were small variations on the roughly 466,000 total labor hours, which could be attributed to rounding. At any rate, these minor variances within Peraton’s proposal are too minor to sustain any prejudice claim, and accordingly GDIT only mentions them in passing. Grumman Data Sys., 15 F.3d 1044, 1048 (Fed. Cir. 1994) (“[S]mall errors made by the procuring agency are not sufficient grounds for rejecting an entire procurement.” (citing Lockheed, 4 F.3d at 960).

Finally, Peraton argues that Plaintiff has not shown prejudice. The erroneously inflated figure, it claims, originated in a mistaken identification of Peraton’s option year 4 hours as “ hours instead of ,” resulting in an identification of “Peraton’s total hours as being 547,540.5 instead of 466,701.8.” Peraton MJAR at 36–38. Peraton argues that the Agency still found the other three option periods to provide reasonable hour estimates. Id. at 37. Peraton argues that these option periods all had amounts “in line with” the inaccurate option period 4 amount, and GDIT has not shown that, were option period 4 to be corrected, it would have changed the outcome of the evaluation. Id. at 38.

25

However, the question is not whether the individual option period amounts were adequate or similar; the question is whether there was a “substantial chance” that, but for substantial inflation in the total labor hours estimate, GDIT may have won the contract. See Sys. Stud. & Simulation, 22 F.4th at 998 (quoting Bannum, 404 F.3d at 1353); WellPoint, 953 F.3d at 1380 (quoting Off. Design Grp., 951 F.3d at 1374); Bannum, 404 F.3d at 1353. The Court concludes that there was a substantial chance that, but for the error, GDIT would have received the contract. Plaintiff was the second-place bidder and was in the “active zone of consideration” required to show prejudice. Allied Tech. Grp., Inc. v. United States, 649 F.3d 1320, 1326 (quoting Statistica, Inc. v. Christopher, 102 F.3d 1577, 1581 (Fed. Cir. 1996)).

The difference between Peraton’s true proposed total hours amount and the inflated amount in the TEP and Award Memorandum is non-negligible (about 15%). See AR 2963 (citing the lower estimate), 2964 (citing the inflated estimate). The CO’s Award Memorandum concludes that the other bidder in Phase 2, , had proposed an “inadequate” number because its proposed hours posed a potential risk to contract completion. See AR 2956–57. Thus, this reflects that, in the Agency’s view, a sufficiently low hours estimate could affect whether the proposed PWS was a high or low performance risk given the proposed labor hours to accomplish it. Id. Indeed, the Award Memorandum also emphasized Peraton’s low performance risk in its best value determination, so that a change to the performance risk assessment based on potentially insufficient labor hours would have necessarily affected the best value determination. See AR 2975. Therefore, Plaintiff is correct that it was a “qualified bidder and that its proposal would have been improved and its chances of securing the contract increased” if the Agency had completed its technical evaluation and award determination with the accurate, lower hours total for Peraton. Info. Tech. & Applications Corp. v. United States, 316 F.3d 1312, 1319 (Fed. Cir. 2003); see

26

GovCIO, LLC v. United States, 177 Fed. Cl. 579, 593 (2025) (“Plaintiff[] meet[s] the low bar of showing” prejudice). Accordingly, Plaintiff’s MJAR on this ground is granted, and Defendant and Intervenor-Defendant’s MJARs on this ground are denied.

C. Whether CMS Irrationally Evaluated Peraton’s Data Center Solution Plaintiff contends that Peraton’s proposal was noncompliant with the Solicitation’s data center requirement because its data center solution allegedly contravened the Solicitation’s mandate “that vendors propose their own data center.” Pl. MJAR at 21. The Court disagrees: on its face, the terms of the Solicitation permitted Peraton’s proposed solution. In the alternative, even assuming arguendo an ambiguity existed, Plaintiff waived its ability to challenge it post-award.

1. Parties’ Arguments Plaintiff acknowledges that there is nothing improper about “an offeror proposing an alternative solution, as long as their primary one is complaint.” OA Tr. 14:13–15. However, Plaintiff contends the Solicitation plainly barred “an externally hosted cloud-like solution provided ‘as a service.’” Id. at 21, 24. Therefore, Plaintiff asserts, the Agency misevaluated Peraton’s proposal as compliant when it awarded Peraton the contract. Id. at 24. Plaintiff first points to the language of the Solicitation itself. According to the SOO, Appendix A, contractors were to “support development and maintenance of the DME system at their data center.” Pl MJAR at 35–36 (quoting AR803 (emphasis in original)). The SOO uses the term “their,” not “a” or “any,” Plaintiff argues. Pl. MJAR at 21. By outsourcing data center hosting to a third party, Plaintiff asserts that Peraton’s solution is not “‘theirs’, meaning ‘of or belonging to them.’” Id. at 24 & n.10 (quoting Their, Cambridge Dictionary Online, https://dictionary.cambridge.org/dictionary/english/their (last visited July 25, 2026)). Plaintiff reinforces its argument by citing CMS’s responses to bidders in the Q&A. Id. at 21–22. As

27

Plaintiff puts it, “[m]ultiple vendors requested that CMS loosen the directive and allow vendors to rely on less costly external solutions,” but the Agency did not permit them to. Id. at 21–22. Additionally, in Plaintiff’s view, Peraton’s solution is a “cloud-like solution . . . not much different” from the solutions proposed and rejected in the Q&A. Id. at 23. According to Plaintiff, Peraton, by purportedly disregarding the “their” requirement, reaps the benefit of its alleged non- conformity by being able to propose minimal monthly costs for a third-party service. Id. at 22– 23.

Plaintiff adds that, in its view, the Agency failed to apply an accurate understanding of the Solicitation or consider whether an outsourced solution met the Solicitation’s requirement. Id. at 25. Instead, Plaintiff argues that the TEP and CO merely list the aspect as “file transfer services” providers in their evaluations, without explicitrecognition that this was meant to fulfill the RFQ’s data center requirement. Id. (citing AR 2121, 2968). Thus, it contends that the record purportedly shows no consideration of an allegedly significant and problematic part of Peraton’s proposal. Id. at 25–26.

As an alternative argument, Plaintiff argues that, if under the Solicitation it was permissible to quote an solution provided “ ,” then there was a latent ambiguity in the Solicitation. Id. at 26. According to Plaintiff, interpreting “their data center” as excluding externally-hosted cloud-based solutions is reasonable because it assumes that the plain meaning controlled, that “‘their’ means ‘their’—not someone else’s.” Id. at 27. If that understanding was incorrect, Plaintiff contends, then a latent ambiguity was present. Id. Therefore, Plaintiff argues, under the principle of “contra proferentum, a latent ambiguity in a solicitation will be construed against the government.” Id. at 27–28 (quoting Furniture by Thurston v. United States, 103 Fed. Cl. 505, 518 (2012)).

28

Defendant agrees that “hosting was required to be at the awardee’s data center.” Def.

MJAR at 24. However, it asserts that Peraton’s solution ( ) satisfies that requirement. Id. Defendant argues that Plaintiff unduly narrows the Solicitation’s goals to requiring the awardee to have a data center of its own, while ignoring other stated goals like moving toward standardization of all SSMs and improving system quality and performance. Id. In general, Defendant contends, Peraton’s solution demonstrated an “understanding of CMS[’s] vision, particularly real time adjudication/payment and system consolidation initiatives.” Id. at 25 (quoting AR 2970). Peraton advocates for a similar interpretation: the Solicitation never required that all contractors manage their own data centers “at the prime contract level.” Peraton MJAR at 18. At any rate, it points out, such a strict construction would cut against GDIT, because GDIT does not entirely “own” the data center it proposes; rather, it runs its physical mainframe in a leased data center operated by a third party. 7 Id. at 22 (citing AR 1582; AR 1702). Instead, Peraton interprets “their data center” in context to require contractors to provide a system environment “capable of supporting the . . . DME system” while excluding use of the Agency’s own data center in their proposals. Id. at 19 (quoting AR 720), 21–22 (“CMS sought to spell out that offerors could not use CMS’s Hybrid Cloud Data Center and had to use one that they proposed.” (citing AR 803)). In other words, as Peraton contends, “CMS was clear that offerors could not use CMS’s Hybrid Cloud Data Center . . . and instead had to propose using legacy IBM mainframe systems.” 8 Id. at 24. Peraton’s solution was, under that understanding, a perfectly “permitted mainframe solution,” under the SOO. Id. at 21.

7 Defendant also points out this fact. See Def. MJAR at 29, n. 2.

8 Peraton also pushes back against GDIT’s characterization of its solution as “cloud-like” and equating it with other solutions that CMS declined during the Q&A, such as Google and Azure.

29

In the alternative, Defendant and Peraton argue that any ambiguity was patent, and Plaintiff should have raised concerns with any ambiguity pre-award. Def. MJAR at 30–31; Peraton MJAR at 24. Under Blue & Gold Fleet, L.P. v. United States, Defendant argues, Plaintiff has waived its opportunity to complain about ambiguity. Def. MJAR at 31 (citing 492 F.3d 1308, 1313, 1314 (Fed. Cir. 2007)). Peraton argues that any ambiguity must have been evident during the procurement, because multiple offerors asked for clarification on the data center requirements. Peraton MJAR at 24. As Peraton characterizes it, “CMS was clear that offerors could not use CMS’s Hybrid Cloud Data Center as the DME development and testing environment and could not propose to modernize the DME claims processing system to use a cloud-based solution, and instead had to propose using legacy IBM mainframe systems.” Id. Additionally, Peraton asserts that it was GDIT’s choice not to inquire any further. Id. at 24–25. In support, Defendant invokes Judge Reyna’s concurrence in Per Aarsleff A/S v. United States, 829 F.3d 1303, 1316–17 (Reyna, J., concurring), stating that it was unnecessary to determine whether an error was patent or latent because the offerors knew that there were “unresolved questions.” Def MJAR at 30–31. Likewise here, Defendant states, GDIT knew there were questions about the data center requirement which had been raised, and missed its chance to complain about the clarity of the requirement. Id.

Peraton MJAR at 22–23. Peraton first concurs that “CMS was clear” that it “was not looking to move DME to a modern cloud system.” Id. at 23 (citing AR 771, Question 1). But the solution, it argues, was not a modern, cloud-based system. Id. 23–24. According to Peraton, is “simply a different pricing model” for an IBM-based physical mainframe for DME development and testing. Id. at 22. While traditional mainframe solutions used “physical IBM-based hardware programmed using legacy programming languages,” cloud-based solutions use modern programming languages on virtual machines hosted by servers like Amazon’s, Microsoft’s, or Google’s. Id. at 22–23. In contrast to cloud-based solutions where a provider, such as Google, would dictate the hardware on which the user’s data is processed, Peraton’s uses “dedicated hardware” that manages. Id. at 23 (emphasis in original). Therefore, Peraton argues, its solution is “largely the same” as GDIT’s: “dedicated, physical IBM mainframes,” utilizing experience in “these legacy systems.” Id. at 23, 24.

30

Finally, Peraton argues that GDIT has not shown that it held the belief that would not be allowed, or that it would have altered its proposal to include solutions. Id. at 25.

2. Conclusion

The Court concurs with Defendant and Peraton that the Solicitation did not contain a restriction against data center solutions and that Peraton’s proposed solution was consistent with the terms of the Solicitation. In the alternative, any ambiguity was patent and thus Plaintiff has waived its ability to object to it. “It is blackletter law that a procuring agency may only accept an offer that conforms to the material terms of the solicitation.” Furniture by Thurston, 103 Fed. Cl. at 518 (citing Centech Grp., 554 F.3d at 1037); see Centech Grp., 554 F.3d at 1037 (“To be acceptable, a proposal must represent an offer to provide the exact thing called for in the request for proposals.”). Here, the relevant requirement of the Solicitation is that the awardee “support development and maintenance of the DME system at their data center.” AR 803 (emphasis in original). The record shows that an extensive section of Peraton’s proposal was spent describing its relationship to , services, and its projected benefits for the Agency. See AR 1904–23. The Agency found Peraton’s proposal acceptable. AR 2970–75. Contrary to GDIT’s arguments, the terms of the Solicitation do not bar the solution included in Peraton’s proposal.

“We begin with the plain language of the document.” Banknote Corp. of Am. v. United States, 365 F.3d 1345, 1353 (Fed. Cir. 2004). “The Court must take care to interpret the Solicitation ‘in a manner that harmonizes and gives reasonable meaning to all of its provisions.’” DevTech Sys., Inc. v. United States, 176 Fed. Cl. 297, 314 (2025) (quoting Banknote Corp., 365 F.3d at 1353). Language in the Solicitation “is ambiguous . . . if its language is susceptible to more than one reasonable interpretation.” Per Aarsleff A/S v. United States, 829 F.3d 1303, 1310 (Fed.

31

Cir. 2016) (quoting Banknote Corp., 365 F.3d at 1353). In this case, the Solicitation requires work to be performed at “their data center.” AR803 (emphasis omitted). The Solicitation does not explicitly define “their data center.” See id. Plaintiff argues that “their” can only mean “of or belonging to them,” and that the term excludes a data center contracted out “as a service.” Pl. MJAR at 24. It also asserts that the Q&A answers “left no question” as to this limitation. Id. at 22.

However, “their,” does not appear to be as limited on its face as GDIT asserts. 9 “Their”

may mean “belonging,” but it may also have a broader meaning: it means that “which belongs or relates to them.” Oxford English Dictionary, their (adj. & pron.), March 2026, https://doi.org/10.1093/OED/4972214077 (last visited June 7, 2026) (emphasis added). Indeed, Plaintiff itself cherry-picks a narrower definition of “‘theirs’, meaning ‘of or belonging to them.’” Id. at 24 & n.10 (emphasis added) (quoting Cambridge Dictionary Online, “their”). Contracting with for could certainly fall within the scope of the term “their”—“belonging” to Peraton in that Peraton pays a subscription for the service and retains some control; “relating” to Peraton in that the mainframe service is connected to them via their contract with . It is uncertain from where GDIT gets its proposed, selective interpretation of “their,” which would exclude a leasehold but include its own rented space with owned hardware. See Leasehold, Black’s Law Dictionary (12th ed. 2024). Plaintiff’s assertions that “their” means “their,” does not show the contrary. Pl. MJAR at 27.

Nor do the Q&A responses limit the meaning of “their” to exclude . Many contractors received responses emphasizing simply that Defendant would not allow use of the

9 Indeed, as Peraton points out, GDIT itself proposed a bid using its own physical mainframe, but in a data center space not belonging to them entirely, but which it leases out from another party. Peraton MJAR at 22 (citing AR 1582, 1702).

32

Agency CACHE to fulfill the Solicitation. See AR 771, Questions 10, 12–16, 32–34. Two responses are slightly more specific: first, clarifying that “their” means “the awardee[’]s” data center, and next, rejecting a suggestion for “transitioning the VMS development environment to an AWS cloud environment.” AR 771 (Questions 32 and 17). However, the clarification that “their” means “awardee’s” is not helpful, as it gives no specific definition of what does and does not constitute “belonging.” The refusal to use the “AWS cloud environment” precludes a specific type of DME environment—AWS cloud environments. See AR 771, Question 17. While Plaintiff argues that the Q&A’s bar on cloud-based solutions clearly eliminated the solution, Pl. MJAR at 22, Peraton persuasively argues that is not at all “cloud-based” but is a “different pricing model” for the same physical legacy IBM mainframe, and that to say otherwise is to misinterpret “solicitation and industry definitions.” Peraton MJAR at 22, 23–24. Overall, then, the Q&A emphasizes that the contractors would not rely on the CACHE to fulfill the Solicitation but does not shrink the meaning of “their” to exclude a contractor’s data center with a mainframe provided through a third-party contractor such as from falling under “their” or the “awardee[’]s” data center.

Alternatively, if there is an ambiguity in the Solicitation, Plaintiff has waived it. Under Blue & Gold Fleet, L.P. v. United States, 492 F.3d 1308 (Fed. Cir. 2007), “[i]t is settled law that ‘a party who has the opportunity to object to the terms of a government solicitation ... and fails to do so prior to the close of the bidding process waives its ability to raise the same objection subsequently in a bid protest.’” Newimar, 160 Fed. Cl. at 121–22 (quoting Blue & Gold, 492 F.3d at 1313). In Plaintiff’s case, it could see that other contractors were lodging inquiries about the “their data center” requirement. See, e.g., AR 2939 at Questions 10, 12–16, 33, 34. If “their data center” is ambiguous as to its restrictions, then Defendant’s comparison to Per Aarsleff, 829 F.3d

33

at 1316–17 is apt, as it indicates that multiple questions about the topic should have signaled to Plaintiff, if it held a potentially contrary interpretation, that it should request clarification. See Def MJAR at 30–31; Per Aarsleff, 829 F.3d at 1310 (“The presence of ambiguity is demonstrated by the inquiries received during the solicitation process”), 1317 (Reyna, J., concurring) (“The disappointed bidders in this case knew there were unresolved questions about the solicitation's eligibility requirement prior to submitting their proposals, but they failed to protest that issue until after award. As such, their protests of the eligibility requirement were untimely.”). It is undisputed that Plaintiff never questioned the data center requirement prior to award. See OA Tr. 10:18–25. Therefore, even if Plaintiff is correct that an ambiguity existed in the Solicitation, it has waived its ability to complain about it here because that ambiguity was patent. Blue & Gold, 492 F.3d at 1313; Newimar, 160 Fed. Cl. at 121–22.

D. Whether the Solicitation Accurately Reflected the Agency’s Needs GDIT argues that “the Solicitation did not accurately reflect the [Agency]’s data center requirements.” Pl. MJAR at 35. The Court finds that GDIT has not sufficiently established that the Agency’s needs changed prior to award.

1. The Parties’ Arguments GDIT argues that the Agency’s actions violated the requirement that “a procurement solicitation must accurately reflect the Government’s needs.” Pl. MJAR at 37 (quoting Infrastructure Def. Techs., LLC v. United States, 81 Fed. Cl. 375, 393 (2008)); see FAR 15.206(a) (“When, either before or after receipt of proposals, the Government changes its requirements or terms and conditions, the contracting officer shall amend the solicitation.”). Specifically, GDIT claims that CMS solicited proposals that would use the offeror’s own data center, but in fact intended to have the contract performed at the Agency’s data center. Pl. MJAR at 35–37. In

34

support of this argument, GDIT points to the language of the Solicitation once more, as well as the Declaration. Id. at 36–37 (citing Declaration ¶ 5). Plaintiff argues that CMS was obligated to amend the Solicitation when it knew—weeks or months prior to award according to Plaintiff—that it would accept a solution relying on the CACHE. Id. at 37–38; Pl. Reply at 24. Had the Agency amended the Solicitation to reflect that it intended to allow DME support at its own data center, GDIT claims, GDIT “would have structured its quotation accordingly,” to allow for integration into CACHE, and ultimately cut $16.1 million from its pricing. Pl. MJAR at 38.

Defendant responds by arguing that it properly assessed both Plaintiff and Peraton as fulfilling not only the requirement for hosting to be “at the awardee’s data center,” but also the SOO’s goal of moving toward leveraging advanced technologies. Def. MJAR at 24 (citing AR 1726, 784, 786), 25–26 (citing AR 2972). Peraton, in turn, distinguishes GDIT’s claims from cases where the Court of Federal Claims has found that a solicitation did not accurately represent the agency’s needs, such as DZSP 21, LLC v. United States, 139 Fed. Cl. 110, 117–18 (2018), and MVM, Inc. v. United States, 46 Fed. Cl. 126, 131–32 (2000), as these cases involved concrete acknowledgements or removals of requirements before award. Peraton MJAR at 26. It also compares GDIT’s claims to those in Second Street Holdings LLC v. United States, 162 Fed. Cl. 306 (2022), where the Court rejected a claim that the agency’s needs changed when the record showed only that the agency had considered changing its needs, and argues that GDIT’s claims are even weaker. Id. at 26–27. Invoking Allied Tech. Grp., Inc. v. United States, 94 Fed. Cl. 16 41 (2010), aff’d, 649 F.3d 1320 (Fed. Cir. 2011), Peraton also alludes to a potentially higher burden for GDIT to carry if it is alleging pretext. Id. at 28 n.3. Additionally, Peraton finds fault with GDIT’s reliance on the Declaration, as Contracting Officer’s Representative did not have any authority to make material changes to the contract, and GDIT has not moved to

35

include the Declaration in the record, making it a self-serving statement outside of the AR. Id. at 27–28 (citing AR 3068), 28 n.2.

2. Conclusion

Upon a thorough review of the Administrative Record, it is evident that GDIT has not shown that the Solicitation inaccurately reflected the Agency’s needs. “Generally, agencies have broad discretion to define their own needs.” Broaden v. Dep’t of Transportation, No. 2021-2000, 2021 WL 5353890, at *3 (Fed. Cir. Nov. 17, 2021). GDIT argues that the Agency’s actions violated the requirement that “a procurement solicitation must accurately reflect the Government’s needs,” as described in FAR 15.206(a). Pl. MJAR at 37 (quoting Infrastructure Def. Techs., LLC v. United States, 81 Fed. Cl. 375, 393 (2008)); FAR 15.206(a) (“When, either before or after receipt of proposals, the Government changes its requirements or terms and conditions, the contracting officer shall amend the solicitation.”).

However, the record does not indicate that the Solicitation reflected outdated or inaccurate requirements for the Agency. The Solicitation required that proposals put forward a data center solution that relied on “their data center” and not the Agency’s own data center. AR 803. Then, Peraton proposed a data center solution using their contracted , as well as an alternative of using the Agency’s CACHE data center; even Plaintiff agrees that there is no issue with Peraton proposing an additional, alternative solution. See AR 1726, OA Tr. 14:13–15. Finally, the Agency made its award to Peraton, noting that Peraton and GDIT’s ratings were equally superior, but that Peraton’s lower price would provide the best value for the Agency. See AR 2970–75. Given the “presumption of regularity” afforded to Defendant’s actions, and the supportive facts within the Administrative Record, the Court sees no indication that the Agency’s requirements changed from those described in the Solicitation. See Palantir USG, Inc. v. United States, 904 F.3d 980, 989,

36

995 (Fed. Cir. 2018) (“In deciding these motions, the court considers ‘whether, given all the disputed and undisputed facts, a party has met its burden of proof based on the evidence in the record.’” (quoting A & D Fire Prot., Inc. v. United States, 72 Fed. Cl. 126, 131 (2006)).

Plaintiff’s narrative for this apparently hidden change in requirements is that “[b]y all accounts, the Agency knew that it intended Peraton to use CACHE well before the task order was issued.” Pl. Reply at 24–25. However, the only account supporting this supposition is that of the Declaration, which alleges that “[m]ere days after award, the CMS [C]ontracting [O]fficer representative (‘COR’), , contacted the GDIT DME Senior Program Manager, , to discuss CMS’s decision to transition data center activities to the CACHE.”

Pl. MJAR at 36–37 (citing Declaration ¶ 5). At most, this allegation indicates that the Agency may have determined that its needs or approach changed post-award; it does not directly imply that the Solicitation under which CMS awarded Peraton the contract was inaccurate.

In fact, GDIT’s line of logic—that the Declaration evinces that the Agency’s needs had changed because “the Agency knew that it intended Peraton to use CACHE well before the task order was issued,” Pl. Reply at 24–25—insinuates a pretextual award, though GDIT avoids using the term in its briefing. See, e.g., Pl. MJAR at 24 (“With respect to CMS changing its tune on whether CACHE would be used during performance, the Solicitation misrepresented CMS’s actual needs.”), 38 (“[T]he Agency solicited for DME support at the contractor’s data center, but it in fact intended that the development and maintenance work would occur in the Agency’s data center.”); see also Allied Tech. Grp., 94 Fed. Cl. at 41 (stating that “assertions of pretextual disqualification are tantamount to accusations of bad faith”); infra Discussion II.C.3. Such accusations would require “almost irrefragable,” “clear and convincing” evidence, of which there is none here. Galen Med. Assocs., 369 F.3d at 1330; see infra Discussion II.C.3.

37

Attempting to buttress its claims, GDIT has cited a Government Accountability Office (“GAO”) decision, Hoechst Marion Roussel, Inc., B-279073, which held an agency’s solicitation for pharmaceuticals was purposely inaccurate. See Pl. Reply at 23–25. However, that case is inapposite. In Hoechst Marion Roussel, the record expressly described the agency’s required dosage amounts, and the agency never contested that requirement estimate. Hoechst Marion Roussel, Inc., B-279073, 98-1 CPD ¶ 127, 1998 WL 215633, at *2–3 (Comp. Gen. May 4, 1998). The solicitation in Hoechst Marion Roussel described a much larger, “unrealistic estimate” for required dosages, and consequently was at odds with the record. See id. at *2–3. The record here is absent of any such disparity. Indeed, as in Second St. Holdings, there is insufficient evidence that the Agency’s requirements changed prior to award, and in fact, the Agency has indicated that it has not changed its requirements at and at oral argument confirmed the lack of any change to date. See Second St. Holdings, 162 Fed. Cl. at 327 (highlighting the lack of evidence of a change in requirements and stating, “Nor is the Court persuaded that the mere passage of time alone or intervening market changes during the procurement are sufficient to presume a change of requirements in this case.”); OA Tr. at 47:2–4 (Peraton’s Counsel: “Peraton has not received any information that there is any final decision that the [Agency] has actually changed its requirements.”), 47:14–16 (Court (to Defendant’s Counsel): “Do you agree that there has not been a final decision?” Defendant’s Counsel: “Yes. Yes your Honor.”); see also infra at Discussion II.C.3. An extra-record declaration, which at most shows a post-award consideration of a change which has not yet come to pass, falls woefully short of the standard to overcome the presumption of regularity afforded to an agency’s actions. Impresa, 238 F.3d at 1338. Therefore, GDIT has not carried its burden to show that the Agency’s requirements changed during the procurement.

38

E. Whether CMS Conducted a Disparate Evaluation The Solicitation required contractors to include, under factor 3, a draft Quality Assurance Surveillance Plan (QASP) for the agency to consider in developing its own QASP. AR 703. Plaintiff accuses Defendant of treating it irrationally and disparately from Peraton by crediting Peraton alone for its QASP advantages. Pl. MJAR at 32. The Court disagrees, because Defendant provided a rational basis in the record for CMS to assess the two QASP proposals differently.

1. Parties’ Arguments GDIT’s argument comes in two parts. First, GDIT claims CMS irrationally evaluated Peraton’s QASP by praising its “detailed framework” as including “Quality Control Number timeliness measures,” when Peraton’s QASP did not mention “Quality Control Numbers” at all. Id.; see AR 2973, 1760–64. Even assuming, GDIT says, that the phrase “Quality Control Numbers” means to reference the proposal’s specific metrics for the timeliness of proposed “Question Case Numbers” or “Question Control Numbers” for DME’s Help Desk, GDIT had a “nearly identical approach” to the DME Help Desk. Pl. MJAR at 34. For instance, GDIT contends both proposals adhered to the SOO’s minimum timeframe for the help desk to answer questions (90% of questions answered within three business days, 95% answered within seven days, and 100% answered within 30 days). Id. at 34–35 (citing AR 794 (SOO standards), AR 1762 (Peraton’s proposal), AR 1524 (GDIT’s proposal)). Thus, GDIT argues that its proposed QASP was better than Peraton’s. Id. at 35.

Second, GDIT argues that the Technical Evaluation Panel (TEP)’s description of Peraton’s QASP as a “comprehensive framework for monitoring and evaluating performance objectives,” should have applied equally to its own proposed QASP. Id. at 33 (citing AR 2112). GDIT’s proposal was substantially similar, Plaintiff argues, because it contained a table summary for

39

“DME Performance Measurement Approach” comparable to Peraton’s “Performance Requirements Summary”; both summaries showed the relationship between performance measures and PWS delivery and work requirements. Id. (citing AR 1760–64, 1522–25). Furthermore, Peraton’s proposed “Acceptable Quality Level [(AQL)],” GDIT argues, is indistinguishable from GDIT’s description of minimum standards of quality. Id.; see AR 1760, 1522. Finally, both QASP proposals addressed reporting methods. Id. at 33 (citing AR 1761–64, 1522).

Defendant argues that under the Solicitation, the highest possible confidence rating that either GDIT’s or Peraton’s QASPs could have received was “High Confidence”—and each received that rating. Def. MJAR at 36 (citing AR 2078, 2089). To “subdivide[]” the confidence ratings into positive findings, or to posit that more positive findings meant a “‘higher’ high confidence assessment,” Defendant argues, is not supported by the record. Def. Reply at 9–10. Accordingly, Defendant contends the two QASPs were thus not disparately evaluated at all. Id. Peraton argues CMS’s evaluations were neither disparate nor unreasonable, as the two QASPs are not “substantively indistinguishable.” Peraton MJAR at 39. While GDIT and Peraton both included performance standards, Peraton contends that its own QASP set forth “specific” performance standards and “concrete” “Acceptable Quality Levels (AQLs), as well as an evaluation rating and point system. Id. at 40 (emphasis added). Meanwhile, Peraton argues, GDIT’s QASP was positively “nebulous” despite using more words, as it did not provide such concrete performance benchmarks or evaluation ratings. Id. at 41. As for the reference to “Quality Control Numbers (QCN),” Peraton argues that the TEP referenced the “QCNs” merely as illustrative examples of Peraton’s QASP’s specific metrics. Id. at 44. Peraton concludes that the record supports their QASP’s superior “systematic approach to quality surveillance.” Id. at 44.

40

2. Conclusion

After a review of the record, it is evident that Defendant did not disparately or irrationally evaluate Peraton’s and GDIT’s QASP proposals. “To prevail on a disparate evaluation claim,” a protestor must show that the agency disparately rated features of a proposal that are “‘substantively indistinguishable’ or nearly identical from those contained in other proposals.” Steel Point Sols., LLC v. United States, 180 Fed. Cl. 463, 473 (2026) (quoting Off. Design Grp., 951 F.3d at 1372). Plaintiff has not established disparate treatment here because the two QASP proposals are not substantively indistinguishable, and so do not require equal acknowledgement in the Agency’s evaluation. See id; see also Off. Design Grp., 951 F.3d at 1373 (Concluding that two proposals, one of which “did not address the hardware and software requirements” were substantively distinguishable.). For instance, Peraton’s proposed QASP sets forth specific performance standards and concrete AQLs, ranging from “Unsatisfactory” to “Exceptional,” as well as a rating and point system. See AR 1761 (chart presenting performance standards). Each PWS goal correlated with examples of how specific performance standards would earn certain ratings. Id. (column titled “Acceptable Quality Level (AQL)”). Peraton’s specific metrics could plausibly “add[] value from the Government’s perspective.” Def. MJAR at 37; see Harmonia Holdings Grp. LLC v. United States, 152 Fed. Cl. 97, 108–109 (2021) (“While [Plaintiff] included data quality control measures in its proposal, those measures had distinct differences from [Intervenor’s] proposal. For example, [Intervenor] provided performance metrics for numerous task areas that were not similarly reflected in [Plaintiff’s] proposal. These proposals, therefore, were not identical or nearly so.” (internal citations omitted)); see also Steel Point, 180 Fed. Cl. at 474 (finding two proposals were not substantively indistinguishable when one described an element of its proposal in more detail and highlighted specific employees). As Peraton puts it, its proposed QASP “essentially did the legwork” for CMS, which planned to use the awardee’s QASP as a model to

41

develop its own. Peraton MJAR at 42; AR 703 (“Responses must include a draft [QASP] for consideration by the [Agency] in developing its own QASP.”). GDIT’s proposed QASP, on the other hand, did not propose such a detailed system. Compare AR 1523 with AR 1761. For instance, its “Evaluation,” column, which it claims is substantively indistinguishable from Peraton’s “AQL” column, created no rating and point system, and often only provided examples of what 100% completion would entail. Pl. MJAR at 33; AR 1524. Consistent with this difference, the TEP evaluation recognizes Peraton’s QASP’s “detailed framework with specific performance metrics, including Quality Control Number timeliness measures, demonstrating a systematic approach to performance monitoring and contract compliance,” and likewise the Award Memorandum highlights Peraton’s “detailed framework with specific performance metrics.” AR 2111, 2973. This assessment was clearly rational and warranted; Plaintiff has not shown it was irrational or unfair for CMS to highlight certain elements of Peraton’s QASP and not GDIT’s.

As for GDIT’s argument that the TEP’s and Award Memorandum’s praise of Peraton’s QASP proposal erroneously references Peraton’s “Quality Control Number timeliness measures,” this does not make the award basis irrational. Pl. MJAR at 32; AR 2111, 2973, 1760–64 (Peraton’s proposed QASP). It is clear from both the TEP evaluation and the Award Memorandum that the “Quality Control Number” reference is not the hinge upon which Peraton’s High Confidence rating turns; rather, it is used as an example, “demonstrating a systematic approach to performance monitoring and contract compliance.” AR 2111 (emphasis added), 2973 (same). The overall basis of the Agency’s rating “may reasonably be discerned”; Peraton’s QASP offered a “detailed framework with specific performance metrics.” Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983) (quoting Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 286 (1974)); AR 2973; see Grumman Data Sys. Corp. v. Widnall, 15 F.3d

42

1044, 1048 (Fed. Cir. 1994) (“[S]mall errors made by the procuring agency are not sufficient grounds for rejecting an entire procurement.” (citing Lockheed Missiles & Space Co. v. Bentsen, 4 F.3d 955, 960 (Fed. Cir. 1993))).

Finally, even if the TEP had misevaluated the QASP, which it did not appear to do, GDIT has not shown that such a misevaluation prejudiced it in any way. The record is clear that both Peraton and GDIT received a “High Confidence” rating for factor 3. Peraton MJAR at 44–45. Correcting the alleged error GDIT highlights would remove the “sole positive finding” related to the proposed QASPs from Peraton’s Factor 3 evaluation. Pl. MJAR at 32 (citing AR 2111–12); Peraton Reply at 18 (agreeing that it received the sole positive QASP-related finding). However, the remarks underneath the confidence ratings in the Agency’s evaluation are not the ratings themselves, only “things that added to or detracted from” confidence. AR 2089; see also AR 2974 (marking both GDIT and Peraton as “High (superior)” in confidence). Each bidder had multiple positive findings under its rating. See AR 2109–10, 2111–12. GDIT points to nothing in the record to suggest that either removing one praiseworthy aspect from Peraton’s assessed advantages, or adding a bullet point to GDIT for its QASP, would either change the “High Confidence” rating or change the overall evaluation to warrant selection. Indeed, CO’s Award Memorandum indicates that, even if GDIT had earned slightly higher marks in its technical aspect, it would have made no difference. AR 2975 (“Even if GDIT was superior in technical merit it isn’t enough to warrant a ~ $15 million price premium.”). Consequently, Plaintiff has not shown that “that there was a ‘substantial chance’ it would have received the contract award but for” such a minor error. Sys. Stud. & Simulation, 22 F.4th at 998 (quoting Bannum, 404 F.3d at 1353); see also Off. Design Grp., 951 F.3d at 1374 (“De minimis errors in the procurement process do not justify relief.”). Accordingly, Plaintiff’s MJAR is denied on this basis as well.

43

F. Whether the Agency’s Best Value Determination was Unreasonable The irrational labor evaluation, see supra Discussion I.B.2, causes the best value trade-off analysis to be flawed as well. Plaintiff’s other arguments, including that the Oral Presentation Factor was misevaluated in the best value determination, are unconvincing and invite the Court to inappropriately reweigh the procurement factors and improperly step into the Agency’s shoes.

1. Parties’ Arguments First, Plaintiff argues, the Agency’s various misevaluations of Peraton’s quotation (see supra) purportedly render the “best-value determination erroneous.” Pl. MJAR at 39. These misevaluations, Plaintiff avers, include that CMS “did not recognize that Peraton’s lower price came with a reduced contractor workforce.” Pl. MJAR at 40. This means, Plaintiff claims, that the CO did not recognize or perform a tradeoff assessment of Peraton’s lower price versus its smaller contractor workforce. Id. at 43. This, too, GDIT argues, constitutes a lack of fair comparative assessment of the benefits of the competing quotes. Id.

Plaintiff also argues that the best value determination is erroneous because the Agency “failed to acknowledge GDIT’s strength under the heavily weighted Oral Presentation factor.” Id. at 40. Specifically, the CO stated that she adopted the TEP’s evaluation of the non-price factors as her own in the Award Memorandum and that both Peraton and Plaintiff were “equally superior to ,” and “equally technically [rated] as . . . high confidence.” Id. at 40–41 (citing AR 2972, 2974). However, the TEP’s evaluation highlighted five more positive features for Plaintiff’s presentation than Peraton’s. Id. at 40–41 (citing AR 2101–07). Therefore, Plaintiff concludes, the lack of “recognition of GDIT’s superiority nor any true comparison of the competing quotations” means that the best value tradeoff is “woefully insufficient, even for a FAR subpart 8.4 procurement.” Id. Plaintiff claims that the purportedly insufficient best value determination converts the best value tradeoff into a lowest-price selection. Id. Such a conversion is not de

44

minimis error, Plaintiff asserts, because the Oral Presentation factor was the second most important factor and was weighted “far heavier” than Price. Id. (citing DigiFlight, Inc. v. United States, 165 Fed. Cl. 588, 608 (2023)); Id. at 42 (citing AR 695). Therefore, GDIT maintains that had CMS adequately performed its best value tradeoff, it would have had a substantial chance at award. Id.

Defendant argues that under FAR Subpart 8.4’s looser standards, there was no violation of the “‘fundamental fairness of the procurement process.’” Def. MJAR at 40 (quoting Allied Tech. Grp., 94 Fed. Cl. at 44). Next, Defendant emphasizes that the RFQ and FAR Subpart 8.4 do not require “extensive analysis on every point,” merely confidence assessments of “Low, Some, or High.” Id. at 41 (citing AR 2942). Peraton follows a similar line of logic. Peraton MJAR at 50 (citing Steel Point, 180 Fed. Cl. at 478). It also argues, as Defendant does, that different numbers of individual positives for each non-price factor do not contradict a finding of equal confidence ratings for each factor. Id. at 46–50.

2. Conclusion

As noted, the TEP Report’s and Award Memorandum’s evaluation of Pertaton’s proposed LOE is tainted by Defendant’s irrational use of incorrect figures throughout the analysis. See supra Discussion I.B.2. Under a FAR Subpart 8.4 procurement, an Agency need not detail its assessment rigorously. Steel Point, 180 Fed. Cl. at 478 (“In a FAR Part 8.4 best value procurement, ‘the Agency is neither expected nor required to document every decision it makes in rigorous detail.’” (quoting Integrated Fin. & Acct. Sols., LLC v. United States, 161 Fed. Cl. 475, 496 (2022)). “Further, ‘the Federal Circuit’s precedent makes clear that this Court is to provide the deference to the Agency’s best value tradeoff.’” Id. (quoting Sirius Fed., LLC v. United States, 153 Fed. Cl. 410, 423 (2021)). Under the record here, the Agency generally provided sufficient detail and justification for why it thought that Peraton, with its lower price, provided better value than GDIT: the two had similarly “high” confidence ratings across all three non-price factors, but Peraton had

45

the lower price. AR 2970–75. GDIT’s attempts to quibble with its own oral presentation factor ratings are of no merit. First, not elaborating on GDIT’s oral presentation’s virtues is not equivalent to lacking a coherent explanation of its evaluation. DigiFlight, Inc. v. United States, 167 Fed. Cl. 158, 166 (2023). Here, the record shows CMS gave an individualized assessment of GDIT’s oral presentation. AR 2972. Second, more bullet points describing GDIT’s Oral Presentation “high confidence” rating could not increase GDIT’s maximally-high confidence rating. See AR 2089 (describing the ranking system in the TEP evaluation); AR 2975 (“Peraton’s equivalent confidence to GDIT in Factor 1 [] the most important non-price factor, combined with equivalent confidence . . . in Factor 2 and . . . Factor 3, provides the most advantageous combination . . .”). Even were CMS to grant GDIT’s wish and add a few bullet points of description under its rating, the CO’s Award Memorandum indicates that it probably would not have made a difference in view of the significant price differential. As the Award Memorandum states, “[e]ven if GDIT was superior in technical merit it isn’t enough to warrant a ~$15 million price premium.” See AR 2975. The Agency’s individualized assessment of the proposals, the CO’s emphasis on the similarity of the confidence ratings between the two, combined with a significant price delta, is consistent with a rational best value tradeoff to which “the Federal Circuit’s precedent makes clear that this Court is to provide [] deference.” Steel Point, 180 Fed. Cl. at 478 (quoting Sirius Fed., LLC v. United States, 153 Fed. Cl. 410, 423 (2021)).

However, “[a]n agency can only make a rational best value determination if the findings that underlie the best value determination are themselves rational.” Tesla Lab’ys, Inc. v. United States, 172 Fed. Cl. 505, 536 (2024). Here, as noted, the LOE evaluation in the PWS was tainted by the repeated and confused use of incorrect labor hour estimates and does not reflect a rational basis for the agency decision. See supra Discussion I.B.2. The Agency’s best value decision relied

46

on this faulty evaluation. See AR 2944 (“[T]he CO performed further [price] analysis with the assistance of the TEP, . . . in order to assess the appropriateness of the level of effort and skill mix.”), 2944. Accordingly, Defendant’s best value determination lacked a rational basis.

G. Injunctive Relief and Conclusion GDIT seeks for injunctive relief. Specifically: it moves the Court to “set[] aside the Agency’s arbitrary and capricious task order award; issue a permanent injunction requiring the Agency to validate its requirements and issue a revised Solicitation that accurately reflects the [Agency’s] needs, solicit and evaluate revised quotations, and then conduct a new price/technical tradeoff.” Pl. MJAR at 48. However, “[a]n injunction is generally not appropriate when a less drastic remedy is available.” GovCIO, 177 Fed. Cl. at 596 (quoting IAP World Servs., Inc. v. United States, 152 Fed. Cl. 384, 397 (2021)). Under Rule 52.2, “[i]n any case within its jurisdiction, the [C]ourt, on motion or on its own, may order the remand of appropriate matters to an administrative or executive body or official.” Rule 52.2(a). Remand is “the proper course” when “the record before the agency does not support the agency action, if the agency has not considered all relevant factors, or if the reviewing court simply cannot evaluate the challenged agency action on the basis of the record before it.” GovCIO, 177 Fed. Cl. at 596 (quoting Fla. Power & Light Co. v. Lorion, 470 U.S. 729, 744 (1985)). In this action, given that CMS’s errors relate solely to the evaluation of the labor hours in the PWS, the Court determines that a brief remand to CMS to accurately reevaluate the proposed LOE in Peraton’s proposal, is the correct remedy. Consequently, on remand CMS must also take into account such accurate figures and reconsider its overall best value determination, which, as noted above, was tainted by the use of

47

materially inaccurate data in the Agency’s underlying technical and price evaluation of Peraton’s bid. II. Motion to Complete and Supplement the Administrative Record Contemporaneously with its MJAR, Plaintiff filed a Motion to Complete and Supplement the Administrative Record. Mot. to Suppl. The Court denies this motion in its entirety.

A. Timeliness of the Motion to Complete and Supplement the Record Peraton argues that Plaintiff’s Motion to Complete and Supplement the Administrative Record is untimely. Peraton’s Opp. to Mot. to Suppl. at 1, 4–5. As Peraton sees it, GDIT had three weeks to file its Motion, but waited to file contemporaneously with its MJAR, and thus “assumed the risk” of arguing the MJARs without a supplemented record. Id. at 4 (quoting Noblis MSD, LLC v. United States, 180 Fed. Cl. 667, 686 (2026)). As a result, Peraton argues, GDIT’s Motion aims to get an inappropriate “second chance” at argument on the merits if granted. Id. at 5. Peraton relies upon Noblis MSD, LLC v. United States, a recent Court of Federal Claims case denying a motion to supplement purportedly on similar grounds. Id. at 4–5 (citing Noblis, 180 Fed. Cl. at 686).

The Court disagrees. Peraton cannot point to any rule or precedent requiring Plaintiff to file a motion to complete or supplement the record before filing its own MJAR, and the Court is aware of none. Indeed, the court in Noblis remarked that it “was unable to locate any Rule of this Court or case imposing a strict time limit on when such a motion to supplement or to admit extra- record evidence must be filed.” Id. at 686. 10 Furthermore, the circumstances in Noblis are

10 Although the court in Noblis alluded to a minimum deadline in dicta, it was a deadline of “[a]t least by the time of the defendants’ respective MJARs.” Id. By submitting its motion contemporaneously with its MJAR, Plaintiff has even complied with that suggested deadline. See Pl. MJAR, Motion to Complete and Supplement.

48

markedly different than those in the present action. 180 Fed. Cl. at 686. In Noblis, the court denied a motion to supplement with 120 pages of extra-record material far later in merits briefing than in the present action—after the movant had filed its reply brief. Id. at 685–86. Here, there is simply no prejudice to Peraton: Plaintiff filed its Motion and its opening MJAR contemporaneously, and Peraton and the Agency had ample time to respond (and did so). See Peraton Opp. to Suppl.; Def. Opp. to Suppl. As Peraton was not prejudiced and has cited no rule of this Court or binding precedent requiring the Court to reject Plaintiff’s Motion to Complete or Supplement for untimeliness, the Court finds the Motion timely.

B. Motion to Complete Plaintiff moves the Court to order Defendant to complete the record with material it claims is absent from the record but was considered by the Agency or document Agency consideration when awarding Peraton the contract, including all materials “relating to the Solicitation’s requirement for a contractor-furnished data center,” and “related materials . . . surrounding CMS’s decision to abandon the Solicitation’s data center requirement.” Pl. Mot. to Suppl. at 1. Defendant, however, repeatedly states that there are no documents showing pre-award consideration of the CACHE data center. See Def. Opp. to Suppl. at 5, 6. The Court denies this motion because Plaintiff has not mustered enough clear evidence that such documents exist, were generated or considered during decision-making process, and are missing from the record. See CAN Softtech, Inc. v. United States, 175 Fed. Cl. 74, 77 (2025) (Can Softtech I) (“The completion standard . . . asks only whether the plaintiff has provided ‘clear evidence’ that the agency ‘generated or considered’ information ‘during the procurement and decisionmaking process’ but omitted that information.” (citing Poplar Point RBBR, LLC v. United States, 145 Fed Cl. at 494, 489 (2019)).

49

1. Legal Standard

A motion to complete the administrative record is appropriate where a party seeks to add materials that were considered by the agency in reaching its decision or generated during the decisionmaking process but are absent from the record. See id.; Thales USA, Inc. v. United States, 176 Fed. Cl. 258, 277–78 (2025); Insight Pub. Sector, Inc. v. United States, 157 Fed. Cl. 398, 406 (2021); Smith v. United States, 114 Fed. Cl. 691, 695 (2014), aff’d, 611 F. App’x 1000 (Fed. Cir. 2015). The Administrative Record may appropriately include “[d]ocuments that pre-date or are contemporaneous with the relevant decisions.” Arch Chems., Inc. v. United States, 64 Fed. Cl. 380, 386 (2005). The agency enjoys a presumption of regularity in compiling the AR. CAN Softtech, Inc. v. United States, 174 Fed. Cl. 412, 415 (2024) (Can Softtech II) (“Because the [Agency] is responsible for compiling the administrative record for judicial review in an APA- review case (like a bid protest), and its designation of the record is entitled to a presumption of regularity, the burden is on the opposing party.”) “Because the administrative record is presumptively complete,” it is the movant’s burden to provide “clear evidence of material that was generated or considered by the agency but excluded from the record[.]” Thales USA, 176 Fed. Cl. at 277–78 (alteration in original) (quoting BHB Ltd. P’ship v. United States, 147 Fed. Cl. 226, 229 (2020)); CAN Softtech, 175 Fed. Cl. at 77 (“The completion standard . . . asks only whether the plaintiff has provided ‘clear evidence’ that the agency ‘generated or considered’ information ‘during the procurement and decisionmaking process’ but omitted that information from the record filed in court.” (citation omitted)).

The Court of Federal Claims Rules provide a list of “core documents” to be included “as appropriate” in the Administrative Record. Rules App. C ¶ 22; see Insight, 157 Fed. Cl. at 407. Among them are: “correspondence between the agency and the protester, awardee, or other

50

interested parties relating to the procurement,” “records of any discussions, meetings, or telephone conferences between the agency and the protester, awardee, or other interested parties relating to the procurement,” and documents “relating to any pre- or post-award debriefing.” See Rules App. C ¶ 22. Documents such as these “presumptively qualify for inclusion in the Administrative Record,” when “relevant and appropriate” to the “limited nature of the question before [the Court].” Dyncorp Int’l LLC v. United States, 113 Fed. Cl. 298, 303 (2013). A demonstration of bad faith or improper behavior also warrants completion of the record with pre-decisional, deliberative materials. CAN Softtech, 175 Fed. Cl. at 78 n.2 (2025); see also E. W., Inc. v. United States, 100 Fed. Cl. 53, 57 (2011) (citing Beta Analytics Int’l, Inc. v. United States, 61 Fed. Cl. 223, 226 (2004)) (Upon a proper showing, discovery might be allowed [to seek] information intentionally left out of the record, such as evidence of bias or bad faith.”). Here, Plaintiff disavows that it is claiming bad faith. See OA Tr. at 15:14–18.

2. Parties’ Arguments Plaintiff claims that materials “relating to the Solicitation’s requirement for a contractor-

furnished data center,” and “related materials . . . surrounding CMS’s decision to abandon the Solicitation’s data center requirement” are missing from the record. Pl. Mot. to Suppl. at 1. Plaintiff’s request for materials includes both pre- and post-award materials it argues are essential for the Court to determine whether the “Solicitation . . . accurately reflect[ed] the [Agency’s] data center needs.” Mot. to Suppl. at 4. According to Plaintiff, the rapid “timing of the Agency’s decision [to use the CACHE] points to pre-award planning and communication” giving rise to such materials. Reply ISO Suppl. at 4. In support of its theory, Plaintiff relies on the Declaration to claim that on February 18, 2026, after the award to Peraton, CMS informed GDIT (the incumbent) that it would be transitioning from a contractor data center to the CACHE. See Declaration; Pl. Mot to Suppl. at 5. Plaintiff also argues that the Agency’s discussion of

51

the data center requirement in the Q&A implies the existence of further, pre-award record material on the subject. Reply ISO Suppl. at 3.

Meanwhile, Defendant maintains that “[d]iscussions about using data centers other than Peraton’s data center did not occur until days after contract award.” Def. Opp. to Suppl. at 6, 5 (“All of the documents that relate to CMS’s consideration of whether to transition the awardee’s data center services to the CACHE were prepared and exchanged after the award decision was issued.”). Consistent with this, Defendant insists that there are no missing portions of the AR, and “[n]o pre-award records exist that include a consideration of using a data center other than the awardee’s data center.” Id. at 6, 5 (“No Core Documents Exist to Complete the Record”), 6 (“Contrary to GDIT’s contention, no email messages, correspondence, or other documents were prepared or exchanged between the parties that relate to CMS’s consideration of the use of the CACHE data center prior to the award of the contract to Peraton on February 9, 2026.” (citing AR 3044)). Still, Plaintiff contends that the “evidence suggests otherwise”—specifically, Plaintiff argues that the timing of the post award-change (as described in the Declaration) indicates an implausibly quick post-award change of heart at CMS. Reply ISO Suppl. at 3 (citing Def. Opp. to Mot. to Suppl. at 6).

3. Conclusion

The Court denies Plaintiff’s Motion to Complete. It is Plaintiff’s burden to show “clear evidence” that there are gaps in the Administrative Record. CAN Softtech, 175 Fed. Cl. at 77 (“The completion standard . . . asks only whether the plaintiff has provided ‘clear evidence’ that the agency ‘generated or considered’ information ‘during the procurement and decisionmaking process’ but omitted that information from the record filed in court.” (citation omitted)); Poplar Point, 145 Fed. Cl. at 494 (requiring “clear evidence” to demonstrate gap in administrative record). Plaintiff has not met that standard. Plaintiff’s only “proof” that the record is incomplete is that the

52

Agency’s alleged change of plans so soon post-award supposedly “defies belief.” Reply ISO Suppl. at 4. According to Plaintiff, the Agency awarded Peraton the contract on February 9, 2026; and on February 18, 2026, the Agency’s Contracting Officer’s Representative (COR) allegedly called GDIT to discuss transitioning to the CACHE. Id.; Declaration ¶ 5. The Court notes that the Declaration, a source produced and submitted by Plaintiff with its Original and now-supplanted Complaint, is the central basis for this claim, and has not been confirmed by any other sources. See Declaration. Furthermore, although the Declaration relates a conversation about transitioning to the CACHE, that purported transition has not yet come to fruition, according to Defendant’s and Peraton’s statements to the Court on the record at oral argument. See OA Tr. at 47:2–4 (Peraton’s Counsel: “Peraton has not received any information that there is any final decision that the [Agency] has actually changed its requirements.”), 47:14– 16 (Court: “Do you agree that there has not been a final decision?” Defendant’s Counsel: “Yes. Yes your Honor.”); see Pernix Grp., Inc. v. United States, 121 Fed. Cl. 592, 599–600 (2015) (no final agency action when agency had not decided whether to act on Government Accountability Office report that recommended disqualifying plaintiff from competition).

All in all, Plaintiff has presented little more than speculation that there are missing records reflecting CMS’s decision-making process. The agency’s “designation of an administrative record” merits a “presumption of completeness.” Poplar Point, 145 Fed. Cl. at 494; see CAN Softtech, 174 Fed. Cl. at 415. The Declaration’s allegation that Defendant, over a week after award, called GDIT (as the incumbent) to discuss potentially transitioning to the CACHE, indicates at most a post-award change of mind by the Agency. Plaintiff has not sufficiently shown why an alleged post-award change of mind reveals a gap in the record of documents considered or generated by the agency in its decision making. See Thales USA, 176 Fed. Cl. at 277–78; Aset

53

Partners Corp. v. United States, No. 25-2033C, 2026 WL 2208182, at *8 (Fed. Cl. May 29, 2026) (rejecting a challenge to award that “targets events after contract award” because “the fact remains that the award took place prior to the [] changes”).

Plaintiff also makes a more specific claim of purportedly missing pre-award material from the AR. Namely, Plaintiff argues that it is implausible that there is no material showing internal Agency discussion of the data center requirement; otherwise, the agency responses discussing the data center in the Q&A involved “kneejerk” responses. Reply ISO Suppl. at 3. Again, Plaintiff has not provided any evidence for this proposition other than labeling it implausible and inferring that there must be additional decision-making documentation merely because the Agency answered questions about its own Solicitation. Id. at 6. Such speculation is a far cry from “clear evidence” that any other documentation exists. See CAN Softtech, 175 Fed. Cl. at 77; Rotair Aerospace Corp. v. United States, 167 Fed. Cl. 571, 576 (2023) (Rotair II) (denying motion to complete because protester’s “assertions do not constitute clear evidence” that agency created documents protester sought).

Plaintiff’s theory might hold more water if Plaintiff were arguing that the Court should ignore Defendant’s statements that there are no documents to complete the record because of the existence of bad faith. See CAN Softtech, 175 Fed. Cl. at 78 n.2 (2025); see also E. W., Inc., 100 Fed. Cl. at 57 (citing Beta Analytics, 61 Fed. Cl. at 226). Though at oral argument, however, Plaintiff unequivocally stated that it is not alleging bad faith, see OA Tr. at 15:14–18, Plaintiff’s briefing frequently approaches accusations against Defendant of bad faith and pretext, see, e.g., Mot. to Suppl. at 5 (“[G]iven the timing . . . CMS’s decision . . . could not have been out of the blue.”); Reply ISO Suppl. at 2 (implying that Defendant dishonestly phrased its statement that “discussions . . . ‘did not occur until days after contract award’”), 3 (“[T]he [Agency] insists that

54

it did not consider using CACHE until moments after the task order was executed. . . . All available evidence suggests otherwise.”). Multiple times, Defendant states that it has no documents or records evincing consideration of the CACHE solution prior to the award. Def. Opp. to Mot. to Suppl. at 5 (“No Core Documents Exist to Complete the Record”), 6 (“Discussions about using data centers other than Peraton’s data center did not occur until days after the contract award. No pre-award records exist that include a consideration of using a data center other than the awardee’s data center.”). Insofar as Plaintiff raises an argument for completion based on bad faith or pretext, there is no need to order completion of the record, because there has been no “well grounded” or “strong showing” of bad faith or pretext. L-3 Commc’ns Integrated Sys., L.P. v. United States, 91 Fed. Cl. 347, 354 (2010), amended on reconsideration in part, 98 Fed. Cl. 45 (2011); see Impresa, 238 F.3d at 1338. 11 Overall, Plaintiff has not presented the Court with the requisite “clear evidence” that the record is incomplete and that Defendant is mistaken in its claims that it lacks such record materials. The AR is presumptively complete, and Plaintiff has not met the above-referenced standard to merit a further “completion” of the record. Thales USA, 176 Fed. Cl. at 278; see CAN Softtech, 175 Fed. Cl. at 77.

C. Motion to Supplement Plaintiff also asks that, if the Court denies the motion to complete the record, the Court allow supplementation of the record with “all documents and communications relating to (1) the Solicitation’s requirement for a contactor provided data center, and (2) materials related to CMS’s decision to abandon the requirement and instead rely on a[n] [Agency]-furnished data center,” in order to obtain meaningful judicial review. Mot. to Suppl. at 6. Plaintiff’s Motion is denied.

11 See infra Discussion II.C.3.

55

Plaintiff has not sufficiently shown that its requested documents are necessary for meaningful review of whether CMS awarded the contract to Peraton in accordance with the Solicitation and with a rational basis.

1. Legal Standard

A motion to supplement aims to show the existing record is inadequate for meaningful review, and as a solution, to introduce material not considered by the Agency in its decision- making process. Axiom Res. Mgmt. v. United States, 564 F.3d 1374, 1380 (Fed. Cir. 2009); Zolon PCS II, LLC v. United States, 176 Fed. Cl. 279, 288 (2025). “The purpose of limiting judicial review to the record actually before the agency is to guard against courts using new evidence to ‘convert the “arbitrary and capricious” standard into effectively de novo review.’” AgustaWestland N. Am., Inc. v. United States, 880 F.3d 1326, 1331 (Fed. Cir. 2018) (quoting Axiom Res. Mgmt., 564 F.3d at 1380). Accordingly, “‘supplementation of the record should be limited to cases in which the omission of extra-record evidence precludes effective judicial review.’ . . . Judicial review is ‘effective’ if it is consistent with the APA.” Id. (quoting Axiom Res. Mgmt., 564 F.3d at 1380–81). “Judicial review of agency action focuses on ‘the administrative record, which should be supplemented only if the existing record is insufficient to permit meaningful review consistent with the APA.’” Kayhan Space Corp. v. United States, No. 25-104, 2026 WL 1662209, at *11 (Fed. Cl. June 8, 2026) (quoting Axiom Res. Mgmt., 564 F.3d at 1381). Supplementation is an “extreme measure” only permitted in “limited circumstances, such as where the agency failed to consider relevant factors or where there is some evidence of bad faith or improper behavior by agency officials.” Thales USA, 176 Fed. Cl. at 278. “A party seeking to supplement the administrative record bears the burden of showing why the current

56

record is insufficient for effective judicial review.” LS3, Inc. v. United States, 163 Fed. Cl. 1, 7 (2022) (citing Swift & Staley, Inc. v. United States, 159 Fed. Cl. 494, 505 (2022)).

The Supreme Court has required “a strong showing of bad faith or improper behavior” to supplement the administrative record under the Administrative Procedure Act. Citizens to Pres. Overton Park, 401 U.S. at 420 (“[W]here there are administrative findings that were made at the same time as the decision . . . there must be a strong showing of bad faith or improper behavior before such inquiry may be made.”). Although the Federal Circuit has not addressed the question, the Court of Federal Claims has generally adopted the Citizens to Preserve Overton Park standard and required a well-founded and “strong showing” to supplement the administrative record based on allegations of bad faith. L-3 Commc’ns, 91 Fed. Cl. at 354; see Rotair Aerospace Corp. v. United States, 167 Fed. Cl. 56, 61 (2023) (Rotair I) (requiring “strong showing of bad faith or improper behavior” to supplement based on bad faith claim). In order to “supplement the administrative record, allegations of bias and/or bad faith must be based on ‘hard facts’ and sufficiently well-grounded, and not merely innuendo or suspicion.” Palantir USG, Inc. v. United States, 129 Fed. Cl. 218, 237 (2016), aff’d, 904 F.3d 980 (Fed. Cir. 2018) (citing Madison Servs., Inc. v. United States, 92 Fed. Cl. 120, 130 (2010); Pitney Bowes Gov’t Sols., Inc. v. United States, 93 Fed. Cl. 327, 332 (2010) (“The strong showing must have an evidentiary foundation and not rest merely on counsel’s argument, suspicion, or conjecture.” (citing Madison Servs., Inc. 92 Fed. Cl. at 130–31; see also, e.g., Starry Assocs., Inc. v. United States, 125 Fed. Cl. 613, 622 (2015) (allowing supplementation where declarations described three specific instances indicating official’s bias, email correspondence supported such a claim, and evidence showed official’s continued involvement after his purported recusal). Accordingly, in considering a motion to

57

supplement, this Court must examine whether there are well-grounded allegations, beyond mere suspicion, of bad faith.

2. Parties’ Arguments

Plaintiff contends that all communications relating to “the Solicitation’s requirement for a contractor-provided data center” and CMS’s purported “decision to abandon the requirement” are necessary to obtain meaningful judicial review, even if the Agency did not consider such communications in making its award decision. Pl. Mot to Suppl. at 6. This includes all materials “leading up to and immediately following the award decision,” materials relating to the alleged call between and the CME COR, and any similar communications with Peraton. Id. at 8; Reply ISO Mot. to Suppl. at 6. Plaintiff asserts that the requested information is relevant to “the evolution of the Agency’s needs” and therefore to the question of whether the Solicitation accurately reflected them. Mot. to Suppl. at 6. It relies on the Declaration as evidence that CMS’s data center needs were not as they appeared in the Solicitation. Id. at 8 (citing Declaration), 9. To permit Defendant to insulate these materials from review, Plaintiff contends, would allow any agency to avoid review “merely by waiting to make the critical change after the award and claiming that the relevant documents were matters of contract administration.” Pl. Mot. to Suppl. at 9.

Plaintiff makes another argument: that documents revealing a post-award decision to make large changes to the contract requirements may still evince an irrational or unlawful decision because such a decision allegedly “fundamentally upend[s] the nature of the procurement just days after award,” “flout[s] the very terms under which all offerors competed by making broad changes or abandoning express prohibitions,” and “materially change[s] the task order while the ink [is] still wet.” Reply ISO Suppl. at 5. Plaintiff admonishes the Court that to rule against it would

58

reduce “the entire competition process . . . to a sham and bid protest review nothing more than a rubber-stamp of fictional paperwork.” Id.

Defendant claims that Plaintiff is seeking documents irrelevant to the question of whether the award decision was arbitrary and capricious. Def. Opp. to Mot. to Suppl. at 6–7. It does not logically follow, Defendant argues, that just because CMS later considered the CACHE solution, that its needs changed during the procurement. Id. at 7. Peraton agrees and argues that the record, as it exists, is already sufficient to permit meaningful review because it contains the basis for CMS’s award decision and because Defendant has represented that “no pre-award material exists that indicates the award decision was made for reasons other than those provided in the record.” Peraton’s Opp. to Mot. to Suppl. at 7; see Def. Opp. to Mot. to Suppl. at 6 (“Contrary to GDIT’s contention, no email messages, correspondence, or other documents were prepared or exchanged between the parties that relate to CMS’s consideration of the use of the CACHE data center prior to the award.”). As for post-award materials, Peraton contends that GDIT has not “explained how that information is relevant to CMS’s prior award decision”—especially as GDIT has not claimed pretext or bad faith. Peraton’s Opp. to Mot. to Suppl. at 6, 7; see also OA Tr. at 15:14–18 (The Court: “Are you making a bad faith or pretext argument here?” GDIT Counsel: “That's an easy question to refute. No, Your Honor. We're not making any bad faith or pretext arguments.”).

3. Conclusion

Defendant and Peraton are correct. Plaintiff has not carried its burden to show that supplementation is required for the Court to meaningfully review the award decision.

The Court first addresses Plaintiff’s argument that the purported post-award change in data center hosting “fundamentally upend[s] the nature of the procurement just days after award,” “flout[s] the very terms under which all offerors competed by making broad changes or abandoning express prohibitions,” and “materially change[s] the task order while the ink [is] still

59

wet.” Reply ISO Suppl. at 5. This argument appears similar to a “cardinal change” argument. See AT&T Commc’ns, Inc. v. Wiltel, Inc., 1 F.3d 1201, 1205 (Fed. Cir. 1993) (asking “whether [agency] modifications changed the contract enough to circumvent the statutory requirement of competition” and clarifying that “[t]he cardinal change doctrine” overlaps with the question of “whether the modification is within the scope of the competition conducted to achieve the original contract”). However, Plaintiff is clear in its briefing that it does not intend to challenge any change in performance or post-award contract modification. Pl. Reply at 23–24 (“GDIT’s concern is with the Agency’s actions . . . during the competition and immediately after award . . . Given the timing at issue, it is not credible that CMS only ‘changed its approach post-award.’”); 26–27 (“GDIT is objecting to the Agency’s failure to take action prior to issuing the task order, not since.”). Assuming Plaintiff were to assert a CICA violation or cardinal change-type argument, it did not raise this argument until its Reply brief. See id. at 9, 23. It has therefore waived this argument. Newimar S.A. v. United States, 160 Fed. Cl. 97, 124 (2022), aff’d, No. 2022-1949, 2023 WL 8534614 (Fed. Cir. Dec. 11, 2023) (citing Novosteel SA v. United States, 284 F.3d 1261, 1274 (Fed. Cir. 2002)); see also Kelly-Leppert v. United States, No. 2022-1301, 2022 WL 1447952, at *1 (Fed. Cir. May 9, 2022).

Further, even assuming that Plaintiff had timely raised a cognizable “cardinal change”-type claim, it would not be ripe for the Court to review. See Abacus Tech. Corp. v. United States, 180 Fed. Cl. 437, 446 (2026) (“There is no jurisdiction to hear a case that is not ripe, and it is [the] plaintiff[’s] burden to establish ripeness.” (quoting Morris v. United States, 58 Fed. Cl. 95, 97 (2003)) (citing McNutt v. Gen. Motors Acceptance Corp., 298 U.S. 178, 189 (1936)), aff’d, 392 F.3d 1372 (Fed. Cir. 2004)). Indeed, Plaintiff has not established that CMS has actually decided to move its data center to the CACHE. At the June 2026 oral argument, months after the events

60

described in the Declaration allegedly took place, CMS still had not transitioned to the CACHE. OA Tr. at 47:2–4 (Peraton’s Counsel: “Peraton has not received any information that there is any final decision that the [Agency] has actually changed its requirements.”), 47:14–16 (Court (to Defendant’s Counsel): “Do you agree that there has not been a final decision?” Defendant’s Counsel: “Yes. Yes your Honor.”). At most, even taking the Declaration in the light that Plaintiff suggests, it indicates that, post-award, a representative who did not have authority to bind the agency discussed moving to the CACHE in the future. See Decl. at ¶¶ 5, 7; AR 3068 (stating that the COR has no authority to make changes or commitments relating to or conflicting with the contract). As such, the evidence in the record a whole does not establish the “‘consummation’ of the agency’s decision-making process . . . one by which ‘rights or obligations have been determined.’” Sys. Application & Techs., Inc. v. United States, 691 F.3d 1374, 1384 (Fed. Cir. 2012) (quoting Bennett v. Spear, 520 U.S. 154, 177–78 (1997)); see Pernix, 121 Fed. Cl. at 599–600. Accordingly, the Court will not risk “interference until an administrative decision has been formalized and its effects felt in a concrete way by the challenging parties” Abacus Tech., 180 Fed. Cl. at 445 (2026) (citing Abbott Lab’ys. v. Gardner, 387 U.S. 136, 148– 49 (1967), overruled on other grounds, Califano v. Sanders, 430 U.S. 99 (1977); Shinnecock Indian Nation v. United States, 782 F.3d 1345, 1351 (Fed. Cir. 2015)).

Regarding Plaintiff’s argument that supplementation is necessary to review “the evolution of the Agency’s needs” and whether the Solicitation accurately reflected them, the Court is unconvinced. Mot. to Suppl. at 6. Insofar as Plaintiff requests materials reflecting pre-award discussions or decision making, Defendant has consistently maintained that “no email messages, correspondence, or other documents were prepared or exchanged between the parties that relate to CMS’s consideration of the use of the CACHE data center prior to the award.” Def. Opp. to Mot.

61

to Suppl. at 6. GDIT has not produced evidence in the record, other than an inference drawn from the allegations in its Declaration, to cause the Court to doubt Defendant’s compilation of the AR. See CAN Softtech, Inc. v. United States, 174 Fed. Cl. 412, 415 (2024) (“Because the [Agency] is responsible for compiling the administrative record for judicial review in an APA- review case (like a bid protest), and its designation of the record is entitled to a presumption of regularity, the burden is on the opposing party.”)

As for post-award materials, Plaintiff’s arguments still fail. This Court’s review of CMS’s actions “focuses on ‘the administrative record, which should be supplemented only if the existing record is insufficient to permit meaningful review consistent with the APA.’” Kayhan, 2026 WL 1662209, at *11 (quoting Axiom Res. Mgmt., 564 F.3d at 1381). Plaintiff’s bid protest currently under review centers around the Agency’s “failure to take action prior to issuing the task order, not since.” Pl. Reply at 26–27. The Declaration, a self-serving affidavit, at most establishes a potential change in Agency action more than a week post-award. See AR 3227 (reporting award made on Feb. 9, 2026); Declaration ¶ 5 (alleging that on Feb. 18, 2026, GDIT received a call from Contracting Officer’s Representative Puepke). It is unclear how any post-award information Plaintiff requests would shine light on CMS’s “failure to take action prior to issuing the task order,” to address a Solicitation allegedly unreflective of CMS’s needs. Pl. Reply at 26–27. GDIT has not established how, absent a theory of bad faith or pretext, materials discussing the agency’s needs after the close of the procurement will inform the question of whether the Solicitation accurately described CMS’s needs before the award. And, as noted, Plaintiff has disavowed any effort to make a bad faith or pretext argument. See OA Tr. at 15:14– 18.

62

Again, though GDIT repeatedly disavowed its intention to make a bad faith or pretext argument, it appears to be alleging such pretext in all but name: claiming that GDIT was not selected because the Agency had been eyeing Peraton’s cash-saving CACHE solution before award, despite banning it in the Solicitation. See id., OA Tr. 52:4–12; Pl. MJAR at 44 (“With respect to CMS changing its tune on whether CACHE would be used during performance, the Solicitation misrepresented CMS’s actual needs.”), 38 (“[T]he Agency solicited for DME support at the contractor’s data center, but it in fact intended that the development and maintenance work would occur in the Agency’s data center.”); Pl. Mot. to Suppl. at 5 (“[G]iven the timing . . . CMS’s decision . . . could not have been out of the blue.”); Reply ISO Suppl. at 3 (“[T]he [Agency] insists that it did not consider using CACHE until moments after the task order was executed. . . . All available evidence suggests otherwise.”); see Allied Tech. Grp., Inc. v. United States, 94 Fed. Cl. 16, 41 (2010) (“[A]ssertions of pretextual disqualification are tantamount to accusations of bad faith.” (citing Savantage Fin. Servs., Inc. v. United States, 595 F.3d 1282, 1288 (Fed. Cir. 2010)), aff’d, 649 F.3d 1320 (Fed. Cir. 2011). Even under a theory of bad faith or pretext, however, a request to supplement these documents would be unsupported. Plaintiff’s only evidence is the supposed improbability of the idea that Defendant changed its mind and considered Peraton’s alternative CACHE solution more than a week post-award. However, this series of events is not implausible. The Solicitation contemplates a two-week turnaround for the parties to organize a post-award transition. AR 799. Given the quick turnaround, a compressed timeframe might be expected for any post-award decisions. Meanwhile, the record reflects that both Peraton and GDIT had high quality ratings, but Peraton’s price was the lower price; it is plausible for the Agency to have made its decision based on the lower-priced solution and later consider Peraton’s suggestion to allow CACHE use after all. See AR 2970; see also OA Tr. 14:13–15 (Plaintiff’s

63

counsel stating that there is no issue with “an offeror proposing an alternative solution, as long as their primary one is complaint.”).

Plaintiff’s suspicion and innuendo do not rise to a “well-grounded” or “strong showing” of bad faith justifying supplementation. L-3 Commc’ns, 91 Fed. Cl. at 354; Palantir, 129 Fed. Cl. at 237 (stating that in order to “supplement the administrative record, allegations of bias and/or bad faith must be based on ‘hard facts’ and sufficiently well-grounded, and not merely innuendo or suspicion.”); Pitney Bowes, 93 Fed. Cl. at 332 (“The strong showing must have an evidentiary foundation and not rest merely on counsel’s argument, suspicion, or conjecture.”). Plaintiff has only this allegation of a post-award change of plans to sustain its allegations that the “Solicitation misrepresented CMS’s actual needs.” Pl. MJAR at 44; OA Tr. 55:14–15 (Plaintiff’s Counsel: “If we had the hard facts, we wouldn’t be asking you for supplementation.”). That is not enough to convince the Court that it should wield the “extreme measure” of supplementation of extra-record evidence in its review. Thales USA, 176 Fed. Cl. at 278.

*****

64

CONCLUSION

For the reasons stated above, the Court DENIES Plaintiff’s Motion to Complete and Supplement the Administrative Record (ECF No. 35), GRANTS IN PART and DENIES IN PART Defendant’s Motion for Judgment on the Administrative Record (ECF No. 43), GRANTS IN PART and DENIES IN PART Peraton’s Motion for Judgment on the Administrative Record (ECF No. 41), and GRANTS IN PART and DENIES IN PART Plaintiff’s Motion for Judgment on the Administrative Record (ECF No. 34).

Accordingly, the Court ORDERS the following:

(1) This action is REMANDED, to Department of Health and Human Services (“HHS”)

Centers for Medicare and Medicaid Services (CMS or Agency), for a period of 45 days or until September 24, 2026. On remand, the Agency shall, consistent with this ruling, reevaluate Peraton’s proposed Level of Effort for Solicitation No. 75FCMC26Q0007 (Solicitation) consistent with the accurate numbers of total labor hours and FTEs in Peraton’s proposal. The Agency shall also, consistent with this ruling and with the reevaluated proposed Level of Effort, perform a new best value determination.

(2) Defendant’s counsel shall provide a copy of this Court’s Order, which shall constitute service pursuant to Rule 52.2(b)(2), to the Agency. Defendant shall file a Notice with the Court by August 14, 2026, certifying that it has sent a copy of this Order to the appropriate officials at the Agency;

(3) Upon completion of the Agency’s reevaluation, Defendant shall notify the parties of the results and concurrently provide them a copy of the reevaluation and new best value determination referenced in paragraph (1). The parties shall then confer and file a Joint Status

65

Report advising the Court of what, if any, further proceedings are necessary. This Joint Status Report shall be filed within 10 days of the notice of the remand results or by October 5, 2026, whichever is earlier; and (4) This case shall be STAYED during the remand period until October 5, 2026.

(5) The parties are directed to CONFER and FILE a Notice by August 20, 2026, attaching a proposed public version of this Sealed Memorandum and Order, with any competition-sensitive or otherwise protected information redacted.

IT IS SO ORDERED.

Eleni M. Roumel

ELENI M. ROUMEL

Judge

August 10, 2026 Washington, D.C. 12

12 This Memorandum and Order was filed under seal, in accordance with the Protective Order entered in this case. ECF No. 8. On August 20, 2026, the parties filed a Notice under seal. ECF No. 70 (Notice of Proposed Redactions). In the Notice of Proposed Redactions, the parties proposed multiple redactions upon which they agreed. Notice of Proposed Redactions at 1. Intervenor-Defendant Peraton requested that the Court redact one additional term concerning a specific proposal in the procurement, citing potential competitive harm in this still-open procurement as well as future procurements. Id. at 2. Defendant does not oppose this redaction, but Plaintiff GDIT does. Id. at 1–2. The common law right of public access to judicial records is presumed but “is not absolute.” In re United States, 166 F.4th 1001, 1011 (Fed. Cir. 2026) (citing Nixon v. Warner Commc’ns, Inc., 435 U.S. 589, 598 (1978)). In cases where information is proprietary, id., or “in the bid-protest context . . . if its discovery would cause substantial competitive injury, interfere with the confidentiality of the bidding process, or impair national security,” then the Court may deny public access to that information. Glob. K9 Prot. Grp., LLC v. United States, 175 F.4th 1348, 1356 (Fed. Cir. 2026). Here, Peraton’s proposed redaction is appropriate given the posture of this procurement, which continues to be open, and because Peraton reasonably avers that disclosure during procurement could place it at a competitive disadvantage. See id. The sealed and public versions of this Memorandum and Order are identical, except for this footnote, the publication date, and slight variance in pagination.