Stories from the work.

Chapter one — wiring the engine for a global fintech

I joined a crypto-fintech startup as one of its first employees, hired by a co-founder to start marketing while the whitepaper was still a draft. There was no playbook. There was barely a product. There was a question: how do you reach people at scale before you have a team to do it?

My answer was systems. I became the architect for the whole marketing technology stack — segmentation, email journeys, CRM pipelines, paid media across the major ad platforms, and the attribution to tie it together. I built drip campaigns for the corporate, institutional, and lending teams, and wired the CRM so the right leads landed automatically with the right business-dev person instead of sitting in a queue.

A few dozen people became several hundred. The platform grew to close to two million users. I was part of that from the first marketing email to the last.

Then the company hit a restructuring bankruptcy and most of the staff, me included, were let go. Years of building, ended by a memo. It stung — and it pushed me back to my own projects, which is where the next chapters start. The lesson that stuck: the system outlives the campaign, and sometimes it outlives the company. Build things that compound while they last.

Chapter two — content that crossed dozens of markets

Back on my own sites, I wanted to do something a small team usually can't: publish quality content across dozens of countries at once. Different languages, currencies, formats, search behavior. By hand it's impossible. So I built a pipeline.

The first version was naive — translate and ship. It read like a machine wrote it, because a machine did. So I rebuilt it around localization, not translation: a locale context for each market, verified local facts supplied instead of invented, and enough forced variation that no two market pages were near-duplicates. The manual work dropped by more than half, and the output held up across 80+ markets.

Chapter three — a portfolio that runs without me

Today I run several sites that earn, mostly on autopilot, as one person. Agentic systems shipped and running 24/7, with manual work cut by around 80%. Not because I'm superhuman — because I stopped doing the work and started designing the work. Content pipelines, research agents, automated update cycles. I review the output, not the process. A human gate before anything publishes, and an AI doing the heavy lifting up to that gate.

The discipline behind it isn't a productivity hack. It's the same one I use training four to five days a week — running and strength work. You don't get fit on the one hard session. You get fit on the boring, repeated ones that compound. Systems are identical. The unremarkable pipeline you set up and forget is the one quietly carrying the business a year later.

The infrastructure is the asset. The specific site is just where I'm testing it.

The thread through all three

A global company, a content engine, a solo portfolio — different scales, same belief. Range compounds. Systems beat campaigns. And the future of this work is less about doing it faster and more about building the thing that does it. I've bet my own career on that. If you want it pointed at your problem, that's what I do.