Tech Viaduct: Inside the USDS Plan to Rebuild Government Tech [2025]

Introduction: When Government Tech Goes Wrong, Everyone Feels It

You know that feeling when you call a government agency and get transferred four times, only to be told to send a fax? That's not an accident. It's the accumulated weight of decades of decisions made by people who didn't understand technology, contracting processes that reward the cheapest bid instead of the best work, and bureaucratic systems designed to prevent change rather than enable it.

In 2021, when the Biden administration took office, a small band of tech warriors at the United States Digital Service (USDS) believed they could fix this. They recruited talented engineers, designers, and product managers from Silicon Valley and pushed hard against the grain of federal inertia. They made progress on pockets of government services, but the overall system remained broken. A website that should take six months took two years. A procurement process that should cost

Then came 2025. The Trump administration rebranded USDS as part of DOGE (the Department of Government Efficiency), brought in a different kind of tech force with different priorities, and began dismantling many of the structures that regulated how government contractors operated. It was chaotic, often reckless, and frequently damaging to citizen privacy and service delivery.

But here's the thing: those former USDS leaders realized something unexpected. The Trump administration's willingness to blow up entire systems and rebuild them from scratch? That instinct wasn't wrong. The execution was terrible, but the core idea—that government needed radical change, not incremental tweaks—actually resonated with them.

So they decided to build their own plan. Not to restore the old system, but to genuinely reimagine it. They called it Tech Viaduct, and it represents one of the most serious attempts yet to answer a fundamental question: what would government technology look like if we actually built it right?

This article goes deep into what Tech Viaduct is, why it matters, who's behind it, what it actually proposes to change, and why the stakes are higher than you might think. Whether you care about how government works or you're just tired of broken federal websites, this is the story of people trying to fix what's been broken for decades.

TL; DR

• Tech Viaduct is a comprehensive reform initiative led by former USDS administrators to completely rebuild how the US government delivers digital services to citizens
• The plan was sparked by an ironic realization: while the Trump administration's DOGE initiative was destructive, its willingness to challenge government inertia showed what boldness could accomplish if directed toward better outcomes
• The team is working toward specific deliverables by spring 2025, with the goal of having a complete implementation-ready plan by 2029 if a Democratic president takes office
• The initiative is well-funded and well-connected, with backing from the Searchlight Institute ($1 million budget), Obama administration veterans like Denis Mc Donough, and strategic advisers from across Democratic politics
• The core challenge isn't technical—it's systemic: fixing outdated procurement processes, replacing contractor networks that profit from complexity, and rebuilding federal hiring to reward merit over connections

The chart highlights the systemic issues in government tech procurement, with 'Lowest Bidder Wins' having the highest negative impact. Estimated data.

Who Started This and Why It Matters

Mikey Dickerson isn't your typical tech leader. He's a former Google engineer who became the first administrator of the United States Digital Service when President Obama launched it in 2014. He has the kind of reputation that makes government people nervous and tech people respectful: someone who doesn't tolerate bullshit, understands how to build systems at scale, and has been through enough bureaucratic battles to know exactly where the land mines are.

In early 2025, Dickerson was trying to disappear. He was packing up his DC condo to move to an abandoned astronomical observatory in rural Arizona—not because he loves stargazing, but because he wanted to be as far away from Washington politics as possible. He'd watched the USDS get rebranded and dismantled. He'd seen talented people pushed out and replaced with operatives who had different priorities. He was done.

Then Robby Mook called him. Mook had been Hillary Clinton's campaign manager and is a strategist who thinks in terms of decades and movements, not election cycles. Mook suggested Dickerson meet with Denis Mc Donough, who'd served as Obama's chief of staff and then as Biden's Secretary of Veterans Affairs. These aren't typical tech people—they're political operators who've spent their careers inside the government apparatus.

When the three of them got together and talked about what went wrong, they came to an unusual conclusion. The Trump administration's DOGE initiative was destructive, chaotic, and often harmful. But it had done something important: it proved that the scale of change needed in government was actually achievable. Nobody could say anymore that you couldn't move fast in government if you had the political will.

Dickerson explained it with surprising directness: "When I was there, we were severely outgunned, 200 people running around trying to improve websites. Trump has knocked over all the beehives—the beltway bandits, the contractor industrial complex, the union industrial complex."

In other words: the Obama administration and Biden administration had been trying to fix government technology from the inside, working around entrenched interests and established processes. They made progress in specific areas, but they never fundamentally changed the system. Trump's team, by contrast, had shown that you could challenge all those entrenched interests simultaneously if you were willing to take the political risk.

So they asked themselves a different question: what if we took that willingness to upend everything, but directed it toward actually making government work better for citizens instead of breaking it?

Tech Viaduct's proposals aim to significantly reduce the time required for key government tech processes, potentially halving procurement and hiring timelines. Estimated data based on typical process durations.

The Searchlight Institute and the $1 Million Bet

Here's where the story gets interesting from a political perspective. Dickerson was skeptical at first. The idea seemed ambitious to the point of fantasy. But when Mook came back a few months later with actual funding, things changed.

The Searchlight Institute, a think tank focused on novel Democratic policy initiatives, agreed to fund the project to the tune of $1 million. That's not a massive amount by government standards, but it's enough to assemble a serious team and produce serious work. Searchlight isn't a traditional think tank—it's designed to incubate ideas that could become policy platforms in future Democratic administrations.

This funding mechanism matters because it removes the project from the immediate pressure of electoral politics. It's not being funded by a specific candidate. It's not being funded by wealthy donors looking for immediate return on investment. It's being funded as a long-term play: develop the ideas now, advocate for them through the 2028 primary and general election campaign, and be ready to execute if a Democrat wins.

Dickerson, like Michael Corleone in Godfather III, found himself pulled back in. "Just when I thought I was out, they pull me back in," is essentially what happened. But this time, the mission felt different. This time, he wasn't trying to patch the system. He was helping to design a replacement for it.

The advisory team that formed around Tech Viaduct reads like a who's who of Democratic operatives who actually understand technology:

Denis Mc Donough (former Obama chief of staff, Biden's Secretary of Veterans Affairs): Mc Donough brings executive experience of the highest level. He knows how to move things through government machinery and understands the political constraints that shape what's possible.

Alexander Macgillivray (Biden's deputy CTO): This is the technical credibility on the team. Macgillivray worked inside the Biden administration on technology policy and understands both the aspirations and the constraints of working in federal government.

Marina Nitze (former CTO of the VA): Nitze is probably the person on the team with the most hands-on experience trying to modernize a major federal agency. The VA is massive and sprawling—if you can modernize that, you can modernize anything.

Robby Mook: The political strategist who knows how to move ideas through the Democratic political apparatus and turn them into something that could actually become policy.

This combination is deliberate. It's not just tech people. It's tech people who've actually worked inside government, combined with political operatives who understand how to move ideas through the Democratic Party and into administrations. That combination is rare and potentially very powerful.

What's Actually Broken About Government Tech

Before you can fix something, you have to understand what's actually broken. Tech Viaduct starts with a fundamental diagnosis that most government technology reform efforts get wrong.

The problem isn't that government workers are incompetent. The problem isn't even that individual agencies don't try. The problem is systemic—it's baked into the procurement process, the hiring system, the way contractors profit from complexity, and the regulatory structures that make it easier to do things wrong than to do them right.

Here's how it actually works: when a federal agency needs a new system, it goes through a procurement process. This process is designed to ensure fairness and prevent corruption. That's good in theory. In practice, it means:

The RFP (Request for Proposal) is written in legal language, not technical language. It's hundreds of pages long because every edge case and contingency has to be specified upfront. Nobody can understand it except lawyers and consultants who specialize in federal contracting.

The lowest bidder wins, not the best bidder. This creates perverse incentives. Contractors bid low to win the contract, then change orders and scope creep make the final cost 200% of the initial bid. This is so common it's basically built into the cost estimates.

The contractor's incentive is to make the system work, not to make it good. A system that barely works but passes the requirements is profitable. A system that works great but costs more than the bid is a loss leader.

Change is expensive and slow. If the requirements change—and they always do—you need to go through a formal change request process. This can take months and costs money. So instead of adapting, teams just build workarounds.

Hiring is divorced from actual skill and merit. Federal hiring processes are designed to be objective and fair, which is good. But they can take 6-12 months. So by the time someone gets hired, their choice to work in government might have changed. The best people often get frustrated and leave before they even start.

The union industrial complex protects everyone equally. This might sound positive—who doesn't love worker protection?—but in practice, it means you can't easily move good people into better roles or move underperformers out. Everyone gets the same process, the same protections, the same salary bands.

Contractors profit from complexity. Here's the dirty secret: the more complex a government system is, the more money there is for the contractors maintaining it. There's almost no incentive to simplify. If you simplify too much, you put yourself out of business.

The result of all this is that most major government technology systems are:

• Outdated: Running on technology that's 10-15 years old because upgrading requires going through procurement again
• Expensive: Costing 5-10x what similar systems cost in the private sector
• Slow: Taking 2-3 years to build what startups build in 3-6 months
• Fragile: Breaking regularly because they're so complex and so old
• Siloed: Not talking to each other because integrating systems requires going through procurement again

Tech Viaduct's core insight is that you can't fix these problems by hiring more tech people or being nicer to contractors. You have to fundamentally change the incentive structures.

Tech Viaduct aims to significantly reduce procurement and hiring times while addressing systemic issues like contractor incentives, legacy systems, and data silos. Estimated data based on typical government processes.

The Two-Part Vision: Building and Unbuilding

Tech Viaduct has two distinct objectives, and understanding both is crucial to understanding what makes it different from previous government tech reform efforts.

Part One: Building a Master Plan for Government Tech

The first objective is to design a completely new way that government technology and digital services could work. This isn't abstract theory—it's meant to be signature-ready executive orders and legislative drafts that could be implemented immediately if a Democratic president takes office.

The plan includes several specific components:

Rethinking Procurement: The current procurement process takes 6-18 months and produces requirements that are outdated by the time they're finished. Tech Viaduct proposes reimagining this from scratch. This means possibly moving to agile contracting models where work is done in sprints with frequent feedback and payment tied to actual delivery, not just reporting.

Restructuring Hiring: Federal hiring is slow and rigid. Tech Viaduct wants to create a fast-track hiring process for technical talent that reduces the timeline from 12 months to 4-6 months and allows for actual merit-based evaluation instead of just checking boxes on qualifications.

Creating Merit-Based Advancement: Currently, advancement in government is based largely on tenure and credentials. Tech Viaduct wants to create a system where good performance is actually rewarded and poor performance has consequences. This is harder than it sounds in a unionized environment.

Establishing Real Oversight: Part of what went wrong with DOGE was the lack of oversight. Tech Viaduct wants to rebuild inspector general offices and create real accountability mechanisms that actually work.

Building a Technology Strategy: The federal government doesn't really have a coherent technology strategy. Different agencies use different systems, different processes, different contractors. Tech Viaduct wants to create an overarching strategy that allows for some degree of standardization without eliminating agency autonomy.

Dickerson explained the scale of what they're trying to do: "For 10 years we've had tiny wins here and there but never terraformed the whole ecosystem. What would that look like?"

"Terraformed" is the right word. They're not trying to patch the existing system. They're trying to rebuild it from the foundations up.

Part Two: Undoing What DOGE Did

The second objective is equally important but more politically fraught. Tech Viaduct recognizes that between now and 2029, the Trump administration will have made changes that need to be reversed or mitigated.

Some of these changes are relatively straightforward to undo. Personnel changes can be reversed. Organizational structures can be reorganized. But some are much more complex.

The most significant is what happened with data privacy and information siloing. DOGE made policy changes that broke down privacy barriers between federal databases that had previously been kept separate. This was ostensibly done to identify fraud more efficiently, but it also means that information that was previously siloed is now interconnected.

As Dickerson put it: "That was DOGE's whole schtick from the very beginning. That's going to take years to figure out."

The problem is that re-siloing data that's already been interconnected is vastly harder than preventing it from being interconnected in the first place. It's like trying to unscramble an egg. You have to not only separate the data again, but you have to figure out all the places it's already been used or shared, and potentially notify people whose privacy was violated.

This is a genuine technical and legal challenge. Tech Viaduct is planning for this to be a multi-year project that might require legislative action, depending on what actually happened between 2025 and 2029.

The Timeline and the Bet on 2028

Tech Viaduct operates on a specific timeline that reveals something important about how political operatives think about change.

Now through Spring 2025: Develop the initial recommendations and plan. This is the intensive design phase where the actual details get worked out.

Spring 2025 through Election 2028: Advocacy and refinement. They're going to act like a lobbying group, which is interesting because it's one of the few places where you see former government officials explicitly organizing to turn ideas into policy platforms. They'll work with Democratic primary campaigns to see if candidates will adopt these ideas. They'll refine the plan based on feedback from the political process.

2029 onward: If a Democrat wins the 2028 election, the plan is designed to be executable immediately. No lengthy consensus-building. No waiting for a second term. Day one, if possible.

Here's the catch: Tech Viaduct only works if a Democrat wins in 2028. Dickerson acknowledged this directly, though he did note that it might be possible under "a Mc Cain Republican"—but as he pointed out, that animal appears to be extinct.

This is a genuine gamble. They're building something that only works if a specific outcome happens in an election that's years away. That takes a kind of faith that many political people don't have anymore.

But it also suggests something about how serious they are. They're not trying to build something that works incrementally across administrations. They're trying to build something that can be executed decisively if they get the chance.

Tech Viaduct aims to significantly improve procurement speed, talent retention, and data integration in federal operations. Estimated data shows potential improvements with their proposed solutions.

Why This Is Different From Previous Reform Attempts

Government technology reform isn't new. Every presidential administration has had someone advocating for it. Every CIO has wanted to modernize their agency's systems. So why might Tech Viaduct actually work where previous efforts failed?

There are a few reasons:

It's explicitly political. Previous reform efforts often tried to be non-partisan. They tried to appeal to efficiency and good governance and universal values. Tech Viaduct isn't trying to be non-partisan. It's explicitly Democratic, explicitly tied to a Democratic victory in 2028, and explicitly designed to be executed by a Democratic president. This actually makes it more powerful because it's built around a specific moment of political change rather than trying to move incrementally.

It understands the nature of the opposition. Previous reform efforts often underestimated how entrenched the opposition would be. They thought once you proved that something worked better, people would adopt it. Tech Viaduct understands that the contractor industrial complex, the union bureaucracy, and the existing government structures will actively resist change. They're building political and organizational strategies to overcome that resistance.

It has both technical and political credibility. Tech Viaduct is led by people who've actually worked inside government and inside tech. They're not pure tech idealists who don't understand how government works, and they're not pure political operatives who don't understand technology. This combination is rare and potentially very powerful.

It's being done in advance of an election, not after. Most government technology initiatives get created after an election when there's a new administration. By then, there's usually not enough political capital to do anything radical. Tech Viaduct is being done now, so that if a Democrat wins in 2028, there's already a fully developed plan ready to execute.

It's designed for speed, not consensus. Dickerson said explicitly: "In the next few months, the group plans to devise and test a framework that could be executed immediately in 2029, without any momentum-killing consensus building."

That's unusual. Most government processes are designed around building consensus. Tech Viaduct is designed around executing decisively if there's an opportunity to do so.

The Specific Problems They're Trying to Solve

Let's get concrete about what Tech Viaduct is actually trying to fix. These aren't abstract problems—they're specific, measurable failures that affect how government actually works.

Problem 1: Acquisition and Procurement Speed

The federal procurement process typically takes 12-24 months from RFP to contract award. By contrast, a typical software company can go from concept to shipping in 3-6 months. This massive time differential compounds itself: technologies that are cutting-edge when you start the procurement process are outdated by the time you finish it.

Tech Viaduct proposes moving to a different model where:

• Contractors are selected based on demonstrated capability, not lowest bid
• Work is done in agile sprints with frequent delivery and feedback
• Payment is tied to demonstrated delivery, not just time and materials
• Risk is shared more equally between government and contractor

This is actually how cloud companies work. You pay for what you use, and if they don't deliver, you can switch providers.

Problem 2: Hiring and Retention of Technical Talent

The federal government currently loses most talented engineers within 2-3 years of hiring. The pay is lower than private sector, the work can be frustrating, and advancement is slow. Tech Viaduct wants to:

• Streamline the hiring process to 6 months instead of 12+
• Allow for performance-based pay adjustments (not just time-based)
• Create clear career paths for technical roles
• Make it easier to hire contractors with specific skills for specific periods

Problem 3: Data Silos and Integration

Federal agencies have databases that don't talk to each other. This creates massive inefficiencies: you fill out the same form for three different agencies, each one verifies your information separately, etc. Tech Viaduct wants to:

• Create data standards that allow agencies to share information securely
• Build APIs that let systems talk to each other without compromising privacy
• Create clear governance around what data can be shared and when

Problem 4: Legacy System Modernization

Many federal systems are running on COBOL code from the 1970s. You can't modify it without breaking something. You can't replace it without disrupting millions of users. Tech Viaduct wants to:

• Prioritize the most critical systems for modernization
• Create a clear path for legacy system replacement
• Allow for parallel running of old and new systems during transition
• Build in security from the start instead of bolting it on later

Problem 5: User-Centric Design

Most government technology is built around what the agency wants to collect, not what users want to do. Tech Viaduct wants to:

• Put user research at the center of requirements gathering
• Test systems with actual users before launching
• Build feedback loops so systems can be improved after launch
• Make it easy for people to understand what the government needs from them

Estimated data shows that the largest portion of the $1 million funding is allocated to Research & Development, followed by Team Assembly, highlighting the project's focus on innovation and expertise building.

The Opposition and the Political Realities

Tech Viaduct isn't naive about the fact that its proposals will face serious opposition. That opposition comes from multiple directions:

The Contractor Industrial Complex: There are hundreds of large government contractors whose business model depends on complexity and long timelines. If government procurement becomes fast and efficient, those contractors' competitive advantage disappears. They'll fight this hard.

Federal Employee Unions: The federal workforce is heavily unionized, particularly in the technical domains. Union contracts typically make it harder to remove underperformers and make merit-based pay adjustments. Unions will likely oppose any reforms that change these protections.

Existing Government Leadership: People who've built careers in the current system often have a vested interest in keeping it. They understand how to work the existing processes. A new system would require them to learn new ways of operating.

Congressional Appropriations Committees: Congress appropriates money for government technology projects. They have incentives to send that money to contractors in their districts. A procurement process that focuses on merit instead of geography threatens those incentives.

Tech Viaduct's strategy for dealing with this opposition involves:

Building Political Coalition: Getting Democratic politicians to commit to supporting these reforms means that when a Democrat wins, there's already political support for executing them.

Creating Implementation-Ready Plans: Having detailed, specific plans ready to go means that advocates can point to concrete proposals rather than abstract ideals.

Using Presidential Authority: Many of these changes can be implemented through executive orders, which means they don't require congressional approval. This reduces the veto points where opposition can block change.

Accepting Staged Implementation: Rather than trying to transform everything at once, Tech Viaduct is planning for changes to be implemented in phases, starting with the most critical systems.

The DOGE Question: Learning From What Went Wrong

One of the fascinating paradoxes at the heart of Tech Viaduct is how its inspiration came from DOGE—which is exactly the kind of reckless, destructive government initiative that reformers usually fight against.

DOGE's approach was to bring in people from outside government, give them broad authority, and tell them to eliminate what they saw as waste and inefficiency. The results were mixed at best: some genuinely problematic programs got cut, some important safeguards got dismantled, privacy protections were weakened, and a lot of institutional knowledge was lost.

But underneath all of that, there was something Tech Viaduct's leadership recognized: DOGE proved that you could move fast in government if you had the political will. You didn't have to go through the normal processes. You didn't have to build consensus. You could just execute.

The difference is in the direction of the change. DOGE's philosophy was roughly "destroy and hope something better emerges." Tech Viaduct's philosophy is "destroy the old systems deliberately and replace them with something that's been thoughtfully designed."

Dickerson was explicit about this: "They're not wrong about that [the basic idea that it's too hard to get things done]. But the execution is completely different from what it should be."

That's the lesson Tech Viaduct is taking from DOGE: boldness matters, but so does direction. You can be bold in service of actually improving government, not just dismantling it.

Tech Viaduct's success relies heavily on political sophistication and technical expertise, with a strong need for patience and determination to overcome entrenched interests. Estimated data.

The Role of Individual Leaders: Why Mikey Dickerson Matters

One thing that jumps out about Tech Viaduct is how much it depends on specific people, particularly Mikey Dickerson.

Dickerson isn't a typical Washington figure. He's got a reputation for being blunt to the point of rudeness. He doesn't suffer fools. He's impatient with bureaucratic processes that he thinks are wasteful. He's willing to fight against institutional inertia.

These qualities make him effective inside government tech work because they push back against the default tendency to work within existing systems. But they also make him controversial with people who prefer consensus-building and institutional respect.

The irony is that Dickerson's presence on Tech Viaduct is both a strength and a potential vulnerability. It's a strength because his credibility with both the tech community and with people who've worked inside government is significant. People believe he'll push for real change rather than settling for incremental improvements.

It's a vulnerability because if he becomes too associated with the plan, it becomes a personal project rather than an institutional initiative. And personal projects are harder to sustain through elections and changes in personnel.

That's why the rest of the advisory team matters. Mc Donough, Macgillivray, Nitze, and Mook are all credible figures in their own right. The plan isn't Dickerson's plan—it's a plan that Dickerson is part of, but that's owned by multiple people with different networks and relationships.

What Success Would Actually Look Like

If Tech Viaduct's work leads to actual implementation, what would change?

Within Year One:

• A new CTO's office with clear authority over federal technology strategy
• Fast-track hiring process for technical talent that actually works
• First wave of legacy system modernization projects underway
• New procurement rules for technology contracts implemented

Within Years Two and Three:

• Multiple agencies operating under new acquisition models
• Data sharing agreements in place allowing secure information exchange
• Measurable improvements in citizen satisfaction with government digital services
• Federal IT hiring stabilizing with retention rates actually competitive with private sector

Within Five Years:

• Most critical federal systems modernized or in process
• Privacy protections restored to pre-DOGE levels or better
• Cost per government IT project measurably lower while quality is higher
• Federal workforce includes significantly more talented technical people

The hard part is that "success" in this context is really hard to measure. Is a system that takes 18 months instead of 24 months a success? Technically yes, but it's not transformative. Tech Viaduct is shooting for transformation: not incremental improvement but genuine rethinking of how government technology works.

The Broader Context: Why This Matters Beyond Government Tech

You might be thinking: okay, but why should I care? I don't work for the government. I don't care how federal websites work.

Here's why it matters: if government can figure out how to move fast and build good technology, it affects everything.

Right now, government is a laggard in technology. It's where good ideas go to die. If government became even just competent at technology—not leading edge, but competent—it would change what's possible in public policy.

Imagine a federal government that could actually collect and analyze unemployment data quickly enough that policy makers knew what was happening in real time. Imagine federal disaster relief programs that could process claims and send money in weeks instead of months. Imagine student loan systems that didn't require multiple phone calls and months of bureaucracy.

These things are possible with modern technology. They're not happening not because technology doesn't exist, but because the process of acquiring and implementing technology in government is broken.

Fixing that process wouldn't just make government better—it would make government more capable of responding to crises, more capable of adapting to new conditions, and more capable of actually delivering on policy promises.

Beyond that, there's something about how a country works that depends on having a government that functions. When government systems fail or are slow or are unreliable, it undermines confidence in institutions. It makes people think government is incompetent. Sometimes people are right—but often, it's not incompetence, it's broken systems.

Tech Viaduct is betting that fixing the systems could actually restore some confidence that government can work.

Financing and Sustainability: The Money Question

Tech Viaduct has $1 million in funding from the Searchlight Institute to get started. That's enough to do the work of planning and designing. But actually implementing this stuff? That requires a different level of financing.

Part of Tech Viaduct's plan involves identifying where government is currently spending money on technology inefficiently, then redirecting that money toward more effective initiatives. In theory, better technology should cost less, not more. But in practice, transitioning from the old model to the new model requires investment.

Here's a rough example: let's say an agency is spending

The problem is that the upfront cost looks scary to budget planners, particularly if they're budget planners who are used to the current model where you just keep spending money on the old system indefinitely.

Tech Viaduct's strategy is to identify a few showcase projects where the modernization can happen, demonstrate the ROI, and use that to build support for broader change. This is sensible but also means that real transformation will take time—probably 5-10 years to implement across the federal government.

Potential Pitfalls and Why This Could Still Fail

Tech Viaduct is an ambitious proposal, and ambitious proposals often fail. Here are some of the ways it could go wrong:

Political Change: The plan only works if a Democrat wins in 2028. If the current administration wins re-election, or if a Republican wins with different priorities, Tech Viaduct becomes irrelevant.

Loss of Key Personnel: If Dickerson, Mc Donough, or other key figures leave the project or become unavailable, the institutional knowledge and credibility could suffer.

Inability to Build Coalition: If Tech Viaduct can't convince Democratic candidates and politicians that this is worth fighting for, it becomes an academic exercise rather than actual policy.

Underestimating Opposition: The contractor industrial complex and entrenched government interests are more powerful than they might appear. They could mount effective campaigns against these changes.

Implementation Complexity: Even if a Democratic president wants to implement Tech Viaduct's plans, actually executing them is hard. Federal agencies have cultures and structures that resist change. People who've spent careers working within the old system don't want to learn a new one.

Unintended Consequences: Disrupting systems that are bad can create temporary chaos. If a modernization project goes wrong, it could undermine support for further change.

Comparing Tech Viaduct to Previous Reform Efforts

There have been several major government technology reform initiatives over the past 20 years:

Bush Administration IT Modernization: This focused on consolidating systems and reducing redundancy. It achieved some success but was eventually overwhelmed by legacy systems.

Obama Administration USDS: This created the Digital Service and focused on high-impact projects. It demonstrated that good teams could improve specific systems, but it struggled to change the overall system.

Biden Administration Tech Investments: This focused on grants and funding for modernization. It provided money but didn't change the underlying procurement processes, so the money was often spent on the old model.

Tech Viaduct is different from all of these because it's trying to change the entire system at once rather than optimize within the existing system. That's more ambitious but also riskier.

The Next Three Years: What to Watch

If you want to follow Tech Viaduct's progress, here's what to watch for:

Spring 2025 Recommendations: The first major deliverable will be when Tech Viaduct publishes its initial recommendations. This will tell you whether they're thinking big or settling for incremental improvements.

2025-2026 Democratic Primary: Watch which Democratic primary candidates adopt Tech Viaduct's ideas or something similar. This will tell you whether the ideas are gaining traction.

2026-2028 Policy Development: As Democrats develop their general election platform, watch whether government technology reform gets significant space. This will indicate whether Tech Viaduct's ideas are becoming mainstream Democratic policy.

Post-2028 Election: If a Democrat wins, the real test will be how quickly and thoroughly they implement the plan.

FAQ

What is Tech Viaduct and why was it created?

Tech Viaduct is a comprehensive government technology reform initiative launched by former US Digital Service leaders and Democratic operatives to fundamentally rebuild how the federal government delivers digital services to citizens. It was created after the Trump administration's DOGE initiative demonstrated that bold change in government was possible, leading reformers to ask: what if that boldness were directed toward actually improving citizen services rather than dismantling safeguards? The project is funded by the Searchlight Institute with a $1 million budget and aims to produce a complete implementation-ready plan by 2029 if a Democratic president takes office.

Who leads Tech Viaduct and what are their backgrounds?

Tech Viaduct is led by Mikey Dickerson, the first administrator of the US Digital Service under President Obama, who brings deep technical expertise and government experience. The advisory team includes Denis Mc Donough (former Obama chief of staff and Biden's VA secretary), Alexander Macgillivray (Biden's deputy CTO), Marina Nitze (former CTO of the VA), and Robby Mook (Hillary Clinton's campaign manager). This combination of tech leaders, government veterans, and political operatives gives the project both technical credibility and political sophistication.

What are the main problems Tech Viaduct is trying to solve?

Tech Viaduct addresses systemic failures in how the government acquires and implements technology: procurement processes that take 12-24 months instead of 3-6 months, hiring systems that take a year or more instead of months, contractor incentives that reward complexity instead of efficiency, outdated legacy systems running on decades-old technology, and data silos that prevent agencies from sharing information securely. The core insight is that these problems aren't technical—they're systemic, embedded in processes and incentive structures that actually reward inefficiency.

What is Tech Viaduct's two-part plan?

Tech Viaduct has two distinct objectives: first, creating a master plan to completely redesign how government technology works, including new procurement models, fast-track hiring processes, merit-based advancement, and data standards; and second, developing a strategy to undo or mitigate the damage done by the Trump administration's DOGE initiative, particularly around data privacy and information siloing. The work is designed to produce signature-ready executive orders and legislative drafts ready for immediate implementation if a Democratic president takes office in 2029.

When does Tech Viaduct expect to have its work done?

Tech Viaduct is operating on a specific timeline: initial recommendations by spring 2025, advocacy and refinement through the 2028 election, and implementation-ready plans by 2029 if a Democrat wins the presidency. The project is explicitly betting on a Democratic election victory, though Dickerson notes it might theoretically be possible under a Republican president with sufficient commitment to technology modernization, an animal he says appears to be extinct in current politics.

How will Tech Viaduct address opposition from contractor and government interests?

Tech Viaduct recognizes that its proposals will face opposition from the contractor industrial complex, federal employee unions, existing government leadership, and congressional appropriations committees that benefit from current arrangements. Their strategy involves building political coalition through Democratic campaigns, creating implementation-ready plans that demonstrate concrete proposals, using presidential executive authority to bypass congressional veto points where possible, and implementing changes in phases starting with the most critical systems to build momentum and demonstrate success.

What would success look like if Tech Viaduct's plans are implemented?

Success would include a reformed CTO's office with real authority, a fast-track hiring process that actually works, modernized legacy systems, and new procurement rules that prioritize quality over lowest bidder within the first year. Over three to five years, success would mean multiple agencies operating under new acquisition models, secure data sharing between agencies, measurable improvements in citizen satisfaction with government services, federal IT hiring competitive with private sector, and lower overall costs for better quality technology. The ultimate measure would be government becoming genuinely competent at technology instead of being where good ideas go to die.

Conclusion: The Bet on Better Government Technology

Tech Viaduct represents something unusual in American politics: a long-term bet on systemic change rather than incremental improvement. Mikey Dickerson and his team aren't trying to make government technology 10% better. They're trying to fundamentally rebuild how it works.

The irony is that this ambition was partly inspired by the worst possible source: the Trump administration's DOGE initiative. But sometimes you learn what's possible by watching someone do something recklessly that you'd been told was impossible. DOGE proved that the scale of change Dickerson had always wanted to attempt was achievable if you had the political will.

The question now is whether Tech Viaduct can channel that lesson into something constructive. Not destructive change for its own sake, but deliberate, thoughtful change directed toward actually making government work.

That's a harder problem than just breaking things. It requires technical expertise, political sophistication, patience, and willingness to fight entrenched interests. It requires not just good ideas but implementation plans that account for the reality of how government actually works.

Tech Viaduct seems to understand all of that. The team has the right mix of technical credibility and political experience. The funding is real. The timeline is realistic. The political positioning—doing this work now, before an election, so there's a plan ready if a Democrat wins—is smart.

But nothing is certain. Lots of reform initiatives have failed despite having good ideas and capable people. Politics is uncertain. Implementation is hard. Entrenched interests fight change. Personnel changes. Leadership turns over.

Still, Tech Viaduct represents one of the most serious attempts yet to answer a fundamental question: what would happen if the federal government actually fixed how it builds technology? Not marginally improved it, but fixed it.

If Tech Viaduct's work leads to actual implementation of even 30% of what they're proposing, the federal government would be measurably more effective, more efficient, and more capable of responding to citizens' needs. It would take politicians' capability to deliver on policy promises. It would restore some of the confidence that people have in institutions.

That's not a small thing. And if it works, it could demonstrate that change in government isn't impossible—it just requires boldness, smart planning, and political will.

The next three years will tell us whether Tech Viaduct can deliver on that promise or whether it becomes another ambitious reform initiative that looked good on paper but couldn't overcome the weight of institutional inertia.

My guess? If a Democrat wins in 2028, at least some of Tech Viaduct's ideas will be implemented. They're too sensible, too well-thought-out, and too badly needed to be ignored. Whether they'll be implemented to the extent the team hopes—that's the real question. But the fact that a comprehensive plan exists, ready to go, on day one of a new administration? That changes the game significantly.

Because the default path in government is entropy. Things get slower. Systems get more complex. People get more risk-averse. Tech Viaduct is trying to interrupt that entropy, just once, long enough to build something better.

That's worth paying attention to.

Key Takeaways

• Tech Viaduct is a $1M-funded initiative by former USDS leaders to completely redesign federal government technology and digital services
• The project was inspired partly by DOGE's willingness to challenge government inertia, but directed toward constructive reform rather than destruction
• Core problems being addressed: procurement that takes 12-24 months instead of 3-6, hiring timelines over 12 months, contractor incentives rewarding complexity, and data silos
• The plan has two parts: create a master plan for new government tech systems, and undo privacy/data damage from DOGE's changes
• Success depends entirely on a Democratic presidential victory in 2028, making this a long-term political bet on specific election outcome
• Opposition will come from contractor industrial complex, federal unions, and entrenched government interests, requiring sophisticated political strategy

Related Articles

• UK Police AI Hallucination: How Copilot Fabricated Evidence [2025]
• UK Scraps Digital ID Requirement for Workers [2025]
• Roblox's Age Verification System Catastrophe [2025]
• App Store Age Verification: The New Digital Battleground [2025]
• AI-Generated Non-Consensual Nudity: The Global Regulatory Crisis [2025]
• Grok's Deepfake Crisis: How X's AI Sparked Global Regulatory Backlash [2025]