I spent sixteen years running systems. Then I started building them with AI.

The marketing operations years

I’ve spent sixteen years in B2B marketing operations and analytics at enterprise software companies and startups: Rocket Software, SAP Ariba, Collibra, aPriori, Icertis, and now Five9. The work behind the marketing: martech stacks, campaign execution, data strategy, and the KPI frameworks that connect the demand funnel to pipeline.

At Icertis I led a team of 6 across martech, campaign execution, process optimization, and data strategy. At Collibra I led an Eloqua-to-Marketo migration that more than tripled MQL-to-SQL conversion. At SAP Ariba I ran the martech roadmap for a global organization.

The AI journey

Less than two years ago I wrote my first Python code. I didn’t set out to become a builder. I set out to fix problems that one-size-fits-all vendors couldn’t fix: account research that took days, personalization that didn’t scale, reports nobody had time to write.

It kicked off with a single intelligence dossier. The first version took about a month; two months after that, enterprise sellers were piloting it. By month nine, my AI systems had generated thousands of intelligence reports in production, work that earned a company-wide builder award, the first for building AI workflows. Each solution revealed more challenges to be solved: how to turn manual work into tools and software, and how to leverage AI to scale and automate it.

I’m not an engineer, but I am an operator who understands which problems are worth solving through workflows, tools, and agents.

Now

I’m leading marketing operations and helping drive AI transformation at Five9, where I’ve been since June 2026.

Personal work: The Pod

Outside of work, I run The Pod, an experimental multi-agent lab: Atlas, Sift, and Vigil, a persistent agent team running 24/7 on DigitalOcean, built on Google’s Agent Development Kit. Each agent runs on a different model. Atlas runs on Gemini and coordinates, Sift runs on GLM and researches, Vigil runs on DeepSeek and watches the infrastructure, and I experiment with other models. The team researches, writes, publishes to Bluesky and GitHub on its own schedule, and helps me learn multi-agent systems and AI research firsthand.

I also built a dashboard to run the system: assign work through a shared project tracker, monitor each agent’s status and output, and track what the team costs per day. It’s where I experiment, learn, and see what’s applicable.

What I’m working toward

The writing and the building circle the same set of beliefs. AI keeps changing underneath every plan, so organizations should architect their systems for that: most of the time the answer isn’t more AI, it’s automating the workflow around the AI, with real data in, defined outputs, and clear ownership. The gap between what AI can do and what companies do with it closes when teams start building systems and bridging gaps across the organization instead of leaving the work manual.

I also believe the moat isn’t the model. The durable advantage is the judgment a company codifies, positioning, buyer insight, and plays that today live in a few people’s heads, written down so workflows and agents can run on it. And managing agents takes an operational layer of its own: visibility into status, cost, and output, the same way you’d run any production system.

Put those together and you get the aspiration behind the north star I write about: imagine operating at the speed of strategy. I’m not quite there yet but it’s what I’m passionate about working toward, one learning at a time.

Writing and workshops

I write Applied AI for Marketing Ops on Substack: practical AI implementation for operations professionals, no hype, just what actually works in corporate environments with enterprise constraints. I also lead AI workshops.

Before all of this: Emory University, a BBA in Organization & Management and Business Communications, and a BA in Sociology.

See the systems →Read the writing →