The Office of Personnel Management’s new solicitation to consolidate 119 human resources systems into a single, governmentwide human capital management platform deserves credit for ambition. The intent is sound: one FedRAMP-authorized system of record that finally gives government leaders a unified view of the federal workforce.

But did we learn nothing from the Defense Department’s failed attempt to consolidate its five estates? Seven years, millions of dollars and they ultimately pulled the plug. That effort covered roughly 900,000 records across five agencies, and now we want to take on 2 million records across 119? If we charge ahead as written, this will become another expensive cautionary tale: a colossal waste of time, resources and goodwill.

Meanwhile, HR professionals already stretched thin by reductions in force, budget cuts, deferred resignation program realignments, retirements and the immense overhead of doing HR on legacy systems are expected to field a new enterprise system on top of everything else. It’s not modernization; it’s madness. The only thing that comes to mind when I read this plan is Sgt. Al Powell’s famous line from Die Hard: “Why don’t you wake up and smell what you’re shoveling?”

A vision worth fighting for

OPM Director Scott Kupor’s statement lays out a compelling aspiration: “A single, pan-government core human capital management system that gives real-time visibility into the workforce and drives effective workforce management.” The argument is logical; one system instead of 119 could yield efficiency, consistency and transparency.

But as someone who has spent decades building, fixing, patching and modernizing HR systems across multiple agencies, I can tell you that consolidation of this magnitude is not just an IT project. It’s a wholesale reconstruction of how the federal workforce is modeled, paid and governed.

What about the data?

Nowhere in this plan does it explain how the transition would actually happen.

Migrating 2 million personnel records from 119 disparate systems, many with incomplete, inconsistent or outdated data, isn’t a configuration exercise; it’s a data-engineering campaign of historic proportions. Cleaning, mapping, validating and migrating that data could easily cost more than the new system itself. And that’s just the forward-facing data. The real nightmare lives in the past.

Federal HR systems carry decades of historical data, each era governed by its own rules, pay codes and policy interpretations. Over time, new laws, executive orders and pay authorities have been layered on top, creating generations of records that reflect shifting logic. When a correction is required, every single one of those historical records must be recalculated according to the rules that applied at the time. That’s not a matter of conversion, it’s digital archaeology.

Reconciling the web of Title 5, Title 21, Title 38, Title 42 and dozens of agency-specific pay authorities within a single data model is the part no one wants to talk about. Those aren’t “settings” in a dropdown menu. They’re statutory logic that evolves with every appropriations cycle. This is what has defeated HCM vendors and payroll modernization efforts for decades, and why it will continue to do so without a realistic, data-first transition plan.

The timeline is a hallucination

The solicitation calls for full implementation by July 4, 2027. That’s roughly 18 months from anticipated award to full governmentwide deployment. For comparison, most single-agency migrations — at the departments of Treasury, Homeland Security and Veterans Affairs — take three-to-five years each, and each of those agencies have a complex web that includes multiple vendors, tools and platforms. What are they to do with those? Years of training data, courseware and compliance efforts. Are those contracts going to be cancelled? Are those vendors going to be told they are done?

No commercial vendor — Oracle, Workday SAP or anyone else — has ever executed a transformation of this scale, under a FedRAMP boundary, across 30 agencies, within that timeframe. The July 4th deadline may be politically poetic, but technically, it’s fantasy.

FedRAMP: The gating reality

The RFP makes FedRAMP Moderate authorization a pre-award requirement. That’s not a paperwork detail; it’s the single hardest gate in federal cloud adoption. Both Workday and Oracle are FedRAMP Moderate authorized, but that doesn’t mean their full HCM suites are ready for governmentwide deployment. Every agency implementation will also require a boundary analysis to confirm compliance within its specific environment and risk posture.

FedRAMP isn’t the end game; it’s the starting line. The real challenge is achieving Authority to Operate (ATO) for 119 separate entities, each with its own data, interfaces and risk posture. In today’s compliance climate, getting that many ATOs in 18 months is all but impossible. Even with FedRAMP Moderate in hand, no one is “flipping a switch” on a governmentwide HCM overnight.

The talent shortage nobody’s talking about

Even if the technology were ready, the workforce to deliver it is shrinking, on both sides of the fence.

The integrators who traditionally handle these massive implementations — Deloitte, Accenture, KPMG, IBM and others — are cutting hundreds of cleared engineers and HR IT specialists after DOGE-driven contract consolidations. Some have filed WARN notices. When those benches disappear, the surge capacity needed to mobilize post-award disappears with them. Hiring and clearing replacements takes 6-to-12 months. You can’t spin up a 2-million-user deployment when your labor ecosystem is in retrenchment.

And what about the federal side? Long-term hiring freezes have hollowed out the very HR professionals who would have to lead and govern this transition. When there’s no hiring, HR people retire or move on. The work doesn’t stop; it just migrates to the “haves,” which right now are Customs and Border Protection and Immigration and Customs Enforcement, leaving smaller or under-resourced agencies barely able to execute the basics. In mid-2018 at the State Department, we lost more than half of our HR staff during that freeze. When the faucet was finally turned back on, it took six months just to fill a position because of the backlog and lack of HR practitioners to staff the vacancies. When assistant secretaries pushed their needs to the front of the queue, the result for us was that it took almost two years to bring in one HR IT GS-14.

This time, the hit is even harder. There’s another freeze, and many of the RIF and DRP departures have been HR professionals. The very people who were shown the door are the ones who could have stitched 119 systems into one. Integrators may know the technology, but they don’t know the policies, programs or legislative constraints baked into each agency’s HCM ecosystem, and that institutional knowledge has left the building.

The technology exists (which is debatable). The delivery ecosystem doesn’t and neither does the institutional memory to make it work.

Requirements vs. reality

Appendix A of the solicitation reads like a wish list from every HR executive of the last twenty years. It expects a single SaaS platform to handle:

• Position management, time and attendance, learning management, telework, analytics and predictive dashboards.
• Bi-directional application programming interfaces (APIs) with USA Staffing, USA Performance and USA Learning — What about Monster, Acendre and other TAS platforms? Skillsoft, Cornerstone and other LMS? Shouldn’t we create an open API for all participants? Or are we putting them out of business by directive?
• Real-time payroll synchronization with the Defense Finance and Accounting Service, the National Finance Center and the Administrative Resource Center — What about NFC? Bubble gum and bailing wire is holding them together, and they still use a flat text file (FESI) to exchange data, and have zero position management or enforcement. They are 15 years overdue for modernization; where is the solicitation for that effort? They serve 156 agencies! Modern vendors will have a meltdown trying to interface with NFC, and talk about data quality? “Agency use field,” anyone?
• Role-based access, SCIM-compliant provisioning, and mobile interfaces — Does anyone here understand that HSPD-12 was implemented differently across the entire federal sector? Even within DHS? And no mention of zero trust? Seriously?
• And, for good measure, “agentic AI” to guide employees through self-service tasks — AI cannot bail us out of this. Someone has been listening to the big vendors and their claims that AI fixes everything*.*

That’s 240 functional requirements spanning statutory pay systems, learning, performance and analytics, all supposedly “out-of-the-box.” In the commercial world, those are separate modules, separate contracts and often separate vendors. Of course, the standard comment from all of the big vendors is, say it with me, “The government needs to adjust the way it does business and use commercial best practices*.*” Maybe some of that lobbying money should be spent on streamlining HR rules so we could all use a common platform.

Governance: The missing ingredient

The RFP assumes OPM and the Office of Management and Budget can act as a single enterprise product owner. That has never existed in federal HR IT. Each Chief Financial Officer Act agency has its own chief information officer, its own bargaining obligations and its own statutory exceptions. Without regulatory harmonization before configuration, the concept of “one system” collapses under the weight of 24 different rulebooks. A someone who participated in the Multi Agency Executive Steering Committee (MAESC) at OPM for the last eight years, those agency challenges are real and unless there is legislative relief to accompany this award, it is a non-starter, and OMB pass backs have told the tale for years. Sorry, HR, not enough money to fix your outdated technology.

**A smarter path forward **

We don’t need to abandon the vision; we just need to execute it like professionals who’ve lived through federal modernization before. The technology exists. The path to get there must be rooted in operational truth, not political theater.

1. ** Start with a federal HR data command center.** Before a single line of code is written or license is signed, establish a permanent, cross-agency data command center staffed with federal HR IT and payroll experts; the people who actually understand Title 5, Title 42, Title 38 and the legacy logic that drives them. Their mission: Standardize data definitions, scrub legacy feeds and reconcile historical pay and personnel records once and for all*.* This team should own the “source of truth” layer that every future HCM system connects to. We have a great start with HCIM, and the Shared Services Consortium and HR line of business get revitalized and become mission control for the effort.
2. ** Sequence, don’t synchronize.** Forget the “big bang.” Pick three pilots, one large (say, DHS or VA), one medium (like State or Commerce), and one small (OPM or the National Science Foundation). Use those agencies to prove the data model, payroll interfaces and FedRAMP boundary. Measure twice, migrate once. Once that baseline is hardened, scaling becomes an engineering problem, not an existential one.
3. ** Fund data remediation as a first-class deliverable.** Data cleanup is not overhead; it’s infrastructure, and it must be funded as such. Every modernization dollar should have a companion data dollar. A modern HCM platform sitting on dirty data is like a Ferrari running on kerosene: it might look impressive until you try to drive it.

If we turn AI loose on our existing HR data mess, the results could be catastrophic. This isn’t just a technical concern; it’s an ethical one. Across every domain, agencies are already sounding alarms about algorithmic bias, hallucinations and decision errors caused by poor data hygiene. Now imagine those same problems applied to federal careers where promotions, pay and performance decisions are amplified by flawed data.

If we want to use AI responsibly in federal HR, the data must be right. Otherwise, we’re not modernizing; we’re automating inequity. Launching an AI-enabled HCM platform on a foundation of bad data would only serve to damage careers faster and spawn a whole new category of litigation.

1. ** Treat FedRAMP as the true critical path.** Authorization and boundary stability aren’t administrative hurdles — they’re the foundation of trust. Bake FedRAMP extension and reauthorization into the schedule and cost model. Build the environment before the configuration, not after. We have heard that streamlining is coming with FedRAMP, but we cannot go so fast as to leave ourselves vulnerable. Security is a must and that is why FedRAMP exists.
2. ** Rebuild the delivery ecosystem.** The integrator bench is hollow. We can’t keep pretending there’s an endless supply of cleared HRIT engineers and policy experts waiting to surge. OPM should create a “Federal HR Modernization Consortium,” a shared-labor pool of cleared federal and contractor personnel trained in HR data, FedRAMP compliance and HCM configuration. Let them rotate across agencies, preserving institutional memory and continuity.
3. ** Adopt a core-plus-ecosystem architecture.** One system doesn’t mean one instance. The government needs a federated baseline core, a shared data layer and configuration standard that all agencies plug into, surrounded by interoperable “ecosystem” modules for unique missions. This protects standardization without erasing flexibility, the same hybrid model used successfully in defense enterprise resource planning environments.
4. ** Build once, govern forever.** The transformation shouldn’t end at go-live. Establish a standing governance council that owns the data model, oversees future upgrades and continuously reconciles policy and system behavior. That’s how you stop modernization from reverting to fragmentation. If OPM wants that role, it is suited to leverage the Human Capital Business Reference Model and Human Capital Information Model to measure compliance and accountability to adhere to established standards.

This isn’t radical, it’s responsible. It’s how NASA works, how every serious enterprise approaches missions that matter. We can do this. We have the talent, the brains and the will. But we can’t pretend complexity will yield to a countdown clock. If we rush it, the only thing we’ll launch is a blame game. Real transformation takes respect for the challenge, not shortcuts through it.

For decades, HRIT has been the government’s afterthought, modernized only when crisis demanded it. Now, for once, it’s at the center of attention. That’s progress.

But progress without realism is just noise.

As someone who’s lived inside this machinery, I can tell you: Modernization isn’t about buying software. It’s about reconciling reality, the laws, the data, the workforce and the limits of physics — while remaining operational 24/7.

Let’s make this moment count. The goal is good. The timeline is a hallucination. And the data, if we respect it, will decide whether this becomes a transformation or another chapter in the long federal tradition of optimistic failure.

*Don Bauer recently retired after 28 years of federal service with the Navy, Customs and Border Protection and the Department of Homeland Security, including the last five years as the chief technology officer for the State Department’s Bureau of Global Talent Management. *

Copyright © 2025 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.