Shann Holmberg is leading agent product development at Espressio AI. We’ve worked together for 7 years at Lunar Strategy — 250+ projects, real budgets, real deadlines. He’s not joining from the outside. He built the systems that became Espressio before it had a name. Here’s how he’s thinking about what comes next.
❓ Q: What drew you to Espressio specifically?
A: We built it because we needed it. At Lunar Strategy, we were running marketing for tech companies, managing campaigns at scale, and the tools weren’t keeping up. So we built our own systems. At some point, we looked at what we had and realized it was more useful than most of what was on the market. That became Espressio. I was already here before it had a name.
❓Q: You’ve spent 7 years running marketing systems on real budgets. How does that shape how you build agent products?
A: Zero tolerance for demos that don’t hold up in production. When you’ve run campaigns with real deadlines and real money, you get a very specific sense of where friction lives, and it’s almost never where vendors think it is. Tools that look great in a walkthrough fall apart fast when a real team actually uses them. My job is to make sure Espressio is the one that doesn’t. We lived these problems before we built solutions to them. That’s the part I keep coming back to.
❓Q: What does “leading agent product development” actually mean at Espressio right now?
A: Deciding where agents replace work, not just assist with it, and building the systems around that. There’s a lot of noise about AI agents. Most of it is automation with better branding. What we’re building is different: agents that sit inside marketing workflows and handle the parts that eat time without adding anything strategic. Research, content production, campaign logic, and reporting. The goal is to give marketers back the hours they’re burning on execution, so they can spend time on the calls that actually matter. That’s the version worth building.
❓Q: What’s the specific problem you’re most focused on solving in your first 90 days?
A: Campaign operations. The gap between strategy and execution in most marketing teams is a process problem, not a talent problem. Teams know what they want to do, but they just can’t move fast enough to do it. We’re building the agent layer that closes that gap. In 90 days, I want a system a marketing team can deploy and see results from before the next invoice cycle. That’s the bar we used at Lunar. We’ll see if we can hit it here.
❓Q: What does a well-built AI agent feel like versus a poorly-built one?
A: A poorly built agent makes you manage it. A well-built one you forget about, until you look at what it produced. The difference is whether the team building it understood the workflow well enough to handle the edge cases, or just got the happy path working and shipped it. Most agents break on step three of a real task. The craft is in what happens when the input isn’t clean, the brief changes mid-run, or the output needs a judgment call. That’s where most of the work actually lives.
Espressio AI builds AI agent systems for growth, marketing, and BD teams — content engines, lead gen pipelines, and competitive intel. If your team is spending more time on execution than strategy, that’s the problem we solve. Learn more at espressio.ai
