Sahil Lavingia’s Insights from His 55-Day Stint at Elon Musk’s DOGE Team

Sahil Lavingia, known for founding Gumroad and his early role at Pinterest, recently shared a candid diary about his 55-day tenure as a software engineer on Elon Musk’s Department of Veterans Affairs (VA) DOGE team. This temporary government organization was formed under President Trump’s executive order to identify inefficiencies and reduce waste within federal agencies.

Lavingia’s account sheds light on the complexities of working within a massive government agency with 473,000 employees. Contrary to his expectations, he found strict rules governing layoffs, where factors like seniority and veteran status outweighed performance considerations. This nuanced approach reflects the VA’s unique mission to serve veterans, complicating straightforward cost-cutting measures.

DOGE itself operated more as an advisory body without direct authority, akin to a McKinsey consulting team. The real decisions on layoffs and reforms came from agency heads appointed by the administration. Lavingia noted that DOGE often became the scapegoat for unpopular decisions, a dynamic Elon Musk also criticized publicly.

Despite being a volunteer with no salary, Lavingia was tasked with identifying wasteful contracts and potential layoffs. He also worked on projects that interested him, such as improving the user experience of the VA’s existing large language model (LLM)-based chatbot. His contributions included tools to scan internal documents for key policy terms, analyze contracts using AI, and build organizational charts.

However, Lavingia lamented that he never received approval to deploy projects that could have significantly improved veterans’ services or sped up claims processing. He was able to open source much of his work, but his efforts to modernize the VA’s technology were constrained by bureaucracy and organizational inertia.

He also observed a surprising lack of knowledge-sharing within DOGE, with engineers often starting from scratch rather than building on collective learnings. This lack of centralized documentation and collaboration hindered the team’s efficiency and impact.

Lavingia’s stint ended abruptly after he spoke with a Fast Company reporter about his work. His access was revoked without warning, highlighting the sensitive nature of reform efforts within government agencies.

Despite frustrations, Lavingia concluded that the VA, while slow and meeting-heavy, is functional and not as inefficient as commonly perceived. His experience underscores the challenge of modernizing large government agencies that must balance operational continuity with innovation.

Lessons from Silicon Valley Meets Government

Lavingia’s journey highlights a fundamental dilemma: how can government agencies embrace cutting-edge technology and agile practices without disrupting essential services? His experience suggests that volunteer tech experts entering government face structural and cultural barriers that limit their impact.

For developers and policymakers alike, this story is a reminder that modernizing public sector technology requires more than just innovative tools—it demands deep understanding of organizational dynamics, stakeholder priorities, and regulatory constraints.

Lavingia’s open-sourced projects and AI-driven tools offer a glimpse into what’s possible when tech talent meets government challenges. Yet, his story also calls for more sustainable, integrated approaches to public sector innovation that empower engineers to deliver real-world impact.