What the VA Tech Sprint Taught Me About Innovation in Government

VA AI Tech Sprint Challenge

In the world of government technology, innovation can feel like a paradox—needed urgently, yet mired in legacy systems, procurement hurdles, and policy constraints. That’s why participating in the U.S. Department of Veterans Affairs (VA) Tech Sprint was one of the most energizing and eye-opening projects I’ve ever led.

Over the course of three intense months, I helped lead a team to develop an AI-powered platform to process community care documentation using NLP. We didn’t win the challenge—but we walked away with something even more valuable: a deeper understanding of what real innovation in the public sector takes.

Here are five lessons I took from the experience.

1. Constraints Drive Clarity

Three months. No pre-defined team. A complex, real-world problem.

Working under a tight deadline forced us to cut through the noise and get laser-focused on solving one thing well. We knew we couldn’t build a polished product. But we could architect a prototype that demonstrated how NLP could extract structured data from faxed referral documents—one of the VA’s major pain points.

That time pressure made us more decisive, and paradoxically, more creative.

2. Cross-functional Collaboration is Non-Negotiable

As the director of our innovation lab, I wore multiple hats—solution architect, product owner, team lead, and partnership manager.

I worked with a senior engineer to design the architecture, sourced external technical partners, and facilitated daily collaboration with our ML engineers. I also stayed closely aligned with VA advisors and potential integration stakeholders. The diversity of perspectives made our solution more practical, but it also required active facilitation to keep everyone aligned.

Innovation doesn’t happen in silos—and certainly not in government.

3. Innovation is About Process, Not Just Tech

We built a solid prototype, yes—but what got us noticed was how we approached the challenge.

We documented our architecture clearly, wrote about our assumptions and trade-offs, and packaged our work into a compelling proposal. That attention to process helped us stand out and created artifacts that could be reused for future projects.

Sometimes the most lasting innovations are the systems and workflows you build along the way.

4. Winning Isn’t the Only Win

We didn’t make it to the final round. And while that stung at first, we quickly saw how much we had gained:

• Visibility with federal healthcare leaders

• A reusable NLP platform we could iterate on

• Confidence and credibility for future government-facing projects

Innovation is rarely a one-shot effort. It’s iterative. The VA sprint was a catalyst, not a conclusion.

5. Government Innovation is a Long Game

Tech sprints are exciting, fast-moving, and filled with smart, mission-driven people. But deploying lasting change in government requires endurance.

After the sprint, we had conversations about what it would take to actually integrate a solution like ours into a VA workflow—compliance, data security, procurement, change management. It’s clear that innovation in government doesn’t stop at the prototype; it starts there.

Final Thoughts

Participating in the VA Tech Sprint was a crash course in public sector innovation—and I’d do it again in a heartbeat. It sharpened my ability to lead under pressure, build quickly with purpose, and bridge the technical with the human and organizational.

If you’re considering diving into govtech, my advice is simple: build boldly, collaborate widely, and be ready to play the long game.

Have a civic tech idea or AI prototype you're exploring? I’d love to connect.