Guardrails and Governance: Building a Data-Driven Federal Government

In this episode of Fed Gov Today, Francis Rose welcomes Kshmendra Paul, one of the federal government’s most influential voices in data strategy, to reflect on two decades of evolving federal data policy and implementation. Paul, who has served as the Chief Data Officer at the VA, Program Manager for the Information Sharing Environment (ISE), and in key roles at the Department of Energy and the Department of Justice, shares candid insights into the progress and pitfalls of federal data usage.

Paul’s journey in government began post-9/11, driven by a deep sense of service. As Chief Enterprise Architect at DOJ, he was struck by the siloed nature of information systems. He argues that “Enterprise Architecture” was never meant to be a purely IT or CIO function—it should have served as a tool for enterprise transformation. This early experience shaped his enduring commitment to integrating data sharing, policy, and technology to drive operational improvements across government.

A key focus of the episode is Paul’s tenure at the ISE, which he describes as a “master class in public choice theory.” He explains that many agencies acted in self-interest rather than aligning for the broader public good, making true information sharing a challenge. His work required navigating complex interagency dynamics, balancing federal priorities with state and local equities, and forging partnerships in unconventional areas outside his core expertise.

Paul shares one of his proudest achievements at the VA—creating a “digital twin” of the agency by integrating over 1,000 datasets into an enterprise data model. This system gave leadership a unified, real-time view of operations during the COVID-19 pandemic. The model not only improved decision-making but also produced tangible results, like saving $90 million in six months through smarter acquisitions. These successes, Paul emphasizes, weren’t purely technological—they were underpinned by strong governance and well-designed data policies.

Looking forward, Paul proposes the creation of a Federal Chief Data Analytics Officer role, separate from the Chief Data Officer, with a clear mandate to drive operational decision support. He highlights that while the Evidence Act has helped institutionalize program evaluation, more must be done to embed real-time, data-informed policy analysis into federal management. He sees data analytics as the “killer app” of the Evidence Act era—central to making government decisions that are fast, relevant, and rooted in evidence.

In conclusion, Paul underscores that lasting government reform begins with data and evidence, but it must be stewarded through policy, leadership, and guardrails that encourage sustained adoption. His message is clear: data isn’t just a tool—it’s the foundation of modern governance.