Herman Leybovich founded Compass Strategy Design after watching the same failure repeat itself at one of the most powerful technology companies in the world. The problem was never the platform. It was what happened to the people the moment implementation began.

Herman Leybovich is the kind of person who notices when something is not working and cannot stop thinking about it until he does something about it. He is also a Marine, which partly explains what happened next.

For years, he watched organizations pour genuine resources and genuine ambition into technology that was supposed to transform how they served people, and then watched those people quietly go back to doing things the old way. Not because the technology was bad. Not because the intentions were wrong. But because somewhere between the inspiring vision at the beginning and the finished product at the end, the actual human beings the work was supposed to help had been forgotten. Their daily experience, their real frustrations, their trust, all of it had been treated as secondary to the technical work. And when you treat people as secondary, they respond accordingly.

Leybovich could not accept that this was inevitable. He founded Compass Strategy Design in Leesburg, Virginia with a conviction he has never moved from: that the difference between technology that transforms and technology that disappoints is not the technology. It is whether someone took the time to genuinely understand the people who have to use it, before a single line of code was written.

A Problem Decades in the Making

The pattern Leybovich observed is not unique to government technology, and it did not start with AI. Paul Mulgrew, who joined Compass as Chief Operating Officer after decades in software development and organizational change, has a book he returns to when he wants to make this point: Peopleware, by Tom DeMarco and Timothy Lister, first published in 1987. The book's central argument is stated plainly in its opening pages: the major problems in software work are not technical. They are sociological. The people doing the work, the environment they are in, the relationships they have with each other and with the work itself, these are what determine whether a project succeeds. Not the tools.

That argument is nearly forty years old. Mulgrew finds it as accurate today as the day it was written, which is precisely the point. The technology has changed beyond recognition since 1987. The human dynamics that determine whether technology gets used well have not changed at all. Organizations keep discovering this the expensive way.

A 2024 survey by the National Association of State Chief Information Officers found that ninety percent of state CIOs named recruiting, retaining, and training qualified staff as their primary concern in technology transformation. And yet the people dimension of implementations, the adoption planning, the genuine work of understanding how employees will actually interact with a new system, is routinely addressed in the final weeks of a project, after the technical configuration is already complete. Research on enterprise technology implementations consistently finds failure rates approaching seventy percent, with weak change management and poor attention to human adoption among the most frequently cited causes.

Mulgrew has spent his career building the operational discipline that closes that gap. Leybovich identified the human problem at the vision level. Mulgrew designed the processes that make solving it consistent and repeatable at the delivery level. Together they have built a company where neither dimension gets subordinated to the other.

Compass's empathy-led discovery process begins not with requirements but with relationships. Before anyone configures anything, the team works to understand the humans who will use what they build: their workflows, their frustrations, their actual daily experience of the problems the technology is supposed to solve. That discovery is not a phase to get through before the real work begins. At Compass, it is where the real work begins. And the people on the other end of that discovery are, in Leybovich's view, doing genuinely important public work that deserves to be treated as such. That belief is baked into how the company shows up, not because it is written on a wall somewhere, but because it is why the company exists.

Herman and Paul

What makes Compass's argument coherent is that it is not made by one person. Herman Leybovich brings the founding conviction: that technology fails when the human thread is dropped, and that the solution is to never drop it. Paul Mulgrew brings the operational architecture that makes that conviction repeatable across every project the company takes on.

Mulgrew is a steady, long-view presence at the leadership table. He and his wife Tina have five children and nine grandchildren. He has spent decades in software delivery, agile methodology, and organizational change, at Bloomberg Industry Group, the Bureau of National Affairs, and Computer Sciences Corporation, building the kind of practical wisdom that comes only from having watched both the successes and the failures across a long career in complex work. In conversation, Herman will occasionally pause and tell you that Paul says something better than he does, and then Paul will say it, and he is right.

What Mulgrew brings to the leadership of Compass is a specific and rare combination: deep technical and process credibility alongside a genuine understanding that the technical and the process are never the actual problem. He has lived the Peopleware argument from the inside. He knows what it costs when teams skip the human work. And he has designed an organization that builds the human work into every phase of delivery rather than treating it as something to address at the end, when it is already too late.

What AI Changes and Does Not Change

The arrival of artificial intelligence into the government technology landscape has changed many things about how Compass works and what its clients need. It has not changed the most fundamental question the company was built to answer.

As AI capabilities become integrated into Salesforce and into government workflows more broadly, the risk of repeating the old failure pattern at a new level of complexity increases rather than decreases. AI tools that genuinely help government workers do their jobs better require the same foundational understanding of those jobs that good Salesforce implementations have always required: who the people are, what they actually do, what they need from their technology, and what will make them trust it or distrust it. Agencies that deploy AI without that foundation will produce AI-powered systems that nobody uses, at higher cost and with greater disruption than a failed CRM implementation.

What Compass has built, the empathy-led approach, the discovery process that begins with human understanding, the culture organized around caring about the people on the other side, is not a legacy methodology that AI makes obsolete. It is the methodology that the AI era makes more critical, because the consequences of skipping it scale with the power of the tools deployed.

Leybovich has spoken about the importance of ensuring AI is both usable and useful in the government context: usable meaning that people can actually interact with it, useful meaning that it genuinely helps them accomplish what they came to work to do. The distinction matters because the two are not automatically connected. A system can be sophisticated and still not be something a government employee would choose to open in the morning. Compass's entire existence is organized around closing the gap between those two things.

What They Are Building For

Leybovich describes his mission in terms that extend well beyond implementation. He spent years scaling an organization from two million to over sixty million dollars in revenue while serving public sector clients. The scale of that experience has given him a long view of what government technology transformation actually requires, and a corresponding patience for the slower, more human-centered work that produces durable results rather than impressive demos.

The name Compass is itself a statement of philosophy. A navigational compass points direction. A mathematical compass, the kind used in geometry and design, draws precise figures from a fixed center point. The company's name alludes to both: directional clarity and the precision that comes from returning always to a defined center. That center, for Compass, is the people the work is meant to serve.

In the age of AI, when every technology vendor has a compelling vision and the tools available to make that vision sound plausible in a presentation, the organizations that will actually transform government service are the ones that know how to hold the human thread from the first conversation through to the last day of deployment. Compass was built to be that kind of organization. And in a market where the same failure pattern has been repeating for decades, that may be the most important thing it offers.

Herman Leybovich is the CEO and founder of Compass Strategy Design, and Paul Mulgrew is its Chief Operating Officer. Compass is based in Leesburg, Virginia. This conversation is part of the Humans & AI: Founder Series by HumanCulture.

Sources

National Association of State Chief Information Officers. (2024). 2024 State CIO survey. NASCIO. https://www.nascio.org/publications/

Kreative Core Tech. (2026). 7 reasons why government ERP projects fail. https://kreativecoretech.com/why-government-erp-projects-fail/