“Soon after publication, my access was revoked without warning,” he wrote.
Lavingia’s stint inside the Department of Government Efficiency (DOGE), overseen by Elon Musk under a Trump executive order, paints a picture of chaos, limited authority, and a government surprisingly more efficient than he expected.
Inside the first days of DOGE
Lavingia joined DOGE in March 2025 as a software engineer at the Department of Veterans Affairs (VA), hoping to make a difference through code. As a volunteer, he received no pay. Still, he came in with enthusiasm, bolstered by a background that included supporting Bernie Sanders’ campaign in 2016 and applying to the U.S. Digital Service during Obama’s presidency.
“On Day 1, I was excited,” he wrote. “I got my government ID and learned I’d be advising the Chief of Staff. My salary? $0.”
His first assignment was to analyse over 90,000 federal contracts for waste. He built tools using large language models (LLMs) to flag problematic contracts, extract HR data, and assist with internal reviews tied to layoffs. He even designed dashboards to help VA teams navigate his findings faster.But the honeymoon didn’t last.
Layoffs, limits and lack of coordination
Soon, Lavingia discovered that improving efficiency came with serious constraints. While he was asked to identify “wasteful” contracts and suggest who to lay off, he ran into a rigid structure where decisions were not purely performance-based.
Veteran status and seniority heavily influenced outcomes. Performance? Less so.
He was also disillusioned by DOGE’s internal culture—or lack thereof.
“I wondered why there wasn’t a centralized DOGE software engineering playbook with all of our learnings,” he wrote. “It seemed like every engineer started from scratch.”
Instead of acting as a nimble tech unit, DOGE felt more like “McKinsey volunteers embedded in agencies”, disconnected and decentralised.
Elon Musk’s role and the ‘Fall Guy’ theory
President Trump had launched DOGE on his first day in office to shake up bureaucracy through technology. Elon Musk was its public face, supported by SpaceX’s Steven Davis and a rotating cast of volunteers like Lavingia.
Despite being viewed by some as decision-makers, DOGE had limited authority.
“DOGE had no direct authority. The real decisions came from the agency heads appointed by President Trump, who were wise to let DOGE act as the ‘fall guy’ for unpopular decisions,” Lavingia wrote.
Musk himself echoed this view, telling The Washington Post that DOGE was Washington’s “whipping boy,” blamed for decisions it couldn’t control.
During his brief time, Lavingia worked on AI tools to support the VA’s operations. He developed an internal chatbot nicknamed VAGPT and explored ways to automate and improve user experience in filing disability claims.
One VA engineer reportedly said Lavingia’s work accelerated AI adoption at the department by a year.
But bureaucratic hurdles blocked his prototypes from reaching production. He was never allowed to ship anything live.
“I was never able to get approval to ship anything to production that would actually improve American lives — while also saving money for the American taxpayer,” he wrote.
He was, however, allowed to open-source some of his work, including a tool that scanned documents for terms related to DEI, COVID policies, and WHO collaborations.
The abrupt end came shortly after he spoke with Fast Company about his experience.
“I would say the culture shock is mostly a lot of meetings, not a lot of decisions,” Lavingia told the magazine. “But honestly, it’s kind of fine — because the government works. It’s not as inefficient as I was expecting, to be honest. I was hoping for more easy wins.”
That interview may have sealed his fate.
He never received an official explanation for his removal, and the White House, VA, and Office of Management and Budget declined to comment.
Lavingia’s story underscores a larger issue: can Silicon Valley-style innovation truly reform government systems? Or does the friction between agile tech and slow-moving bureaucracy inevitably grind good intentions down?
He remains proud of the work he managed to complete but regrets leaving before delivering meaningful, visible change.
“In the end, I learned a lot, and got to write some code for the federal government. For that, I’m grateful,” he wrote. “But I’m also disappointed. I didn’t make any progress on improving the UX of veterans’ filing disability claims or automating/speeding up claims processing, like I had hoped to when I started.”
