Developer turned marketer. Now building the layer in between.

I came up writing code, and it shows. When I landed in digital strategy at alliantgroup in 2012, I looked at the marketing org the way an engineer looks at a slow build: too many manual steps, no feedback loops, everything done by hand that a machine could do better. So I started automating — first reporting, then lead management, and eventually an internal lead scoring platform built alongside the Data Science team, years before 'MarTech stack' was a phrase anyone put on a slide.

That instinct — see the repetitive work, build the system that eats it — turned out to be the whole career.

Across alliantgroup, SnapEngage, Sovrn, and Nylas, my job title said marketing but my output was infrastructure: lifecycle programs, scoring models, routing logic, attribution frameworks, experimentation pipelines. The wins I'm proudest of are systems wins — an 80% lift in free-to-paid conversion at Nylas didn't come from a clever ad, it came from behavior-based lifecycle architecture that treated every user signal as an input to a machine that never sleeps.

Somewhere along the way I noticed the pattern in what I kept getting hired to fix: marketing teams don't fail for lack of ideas. They fail because their ideas are trapped behind manual processes, disconnected tools, and engineering queues. The teams that win are the ones where the distance between 'we should try this' and 'it's live' is measured in hours.

When LLMs crossed the capability threshold, I recognized the moment immediately — this was the automation instinct with a nervous system attached. I didn't add 'AI' to my LinkedIn and call it a day. I built products: AI View Sync, a platform that audits how AI models recommend brands and pushes fixes to production. Heynet, a roster of AI employees that screen calls, appraise resale finds, and draft work on top of a directory of 55,000 locally-enriched company profiles. An AI analyst that reads Google Ads accounts and outputs ranked experiments. A content engine with a human exactly where judgment lives.

I work daily in Claude, Cursor, and MCP servers — this site was built with Claude Code. Prompt engineering, to me, is the new management skill: you're writing job descriptions for a workforce that executes instantly and never forgets a process doc.

The role I play best sits on the seam between marketing and engineering. I can sit with a CRO and argue pipeline math, then open a terminal and ship the fix we just agreed on. I've run the meetings where attribution models go to die, and I've written the code that made everyone finally trust the same number.

That bridge used to be a nice-to-have. In the AI era it's the job. The next generation of great marketing teams will be small groups of sharp people directing fleets of AI systems — and they'll be led by people who can build, brief, and debug those systems themselves. That's the team I want to build. That's what I do.