My first actual business was a handyman operation around the neighborhood when I was a kid. Mowing lawns, cleaning gutters, hauling stuff to the curb, small repair jobs, anything a neighbor would pay a middle-schooler to do. I printed flyers. I kept a notebook of who had hired me for what, and when to follow up for a repeat job. I set rates that were fair but not the lowest, because I figured out pretty fast that the lowest bidder got treated like the lowest bidder.
What I remember is not any particular job. It is the feeling of watching a system I had designed actually work. Someone had a need, I had the solution, and the loop closed in front of me every Saturday morning.
That has been the through-line of everything I have done since. I am not a marketer who learned to build software. I am a builder who happened to land in marketing long enough to see the specific shape of the problems there. The software part came later. The instinct was there the whole time.
The Through-Line
If I zoom out on my resume the way you would zoom out on somebody else's, the pattern is almost embarrassing.
I ran a social media network in my early twenties that ended up with over ten million followers across a bunch of niche accounts. I was not a social media expert, I just kept seeing that the tools the big publishers used were too expensive and too clunky, and I kept building small scrappy systems that did the same thing with a Google Sheet and a cron job. Every time something got easier for me, I tried to turn that into leverage, more accounts, more posts, more content hubs siphoning traffic from the network.
In my late twenties I co-founded a direct-to-consumer e-commerce brand and ran it for five years until the acquisition. I was the organic-growth half. We got to hundreds of thousands of sessions a month with zero paid, because every time I ran into a process that did not scale (product content, email sequencing, merchandising) I would figure out how to turn it into a repeatable system before I hired someone to run it. The hires came after the system, not before.
I did intelligence analysis work early on too, that is the part of my background that looks like the odd one out but is actually the core of everything. A summer internship in the Intelligence Division at the Nassau County PD, and analyst work for the Society of Former Special Agents of the FBI. What I learned in those rooms was how to collect, correlate, and make sense of signal in messy data. That is what marketing engineering is, most days. You are not writing fancy algorithms. You are figuring out which three inputs actually matter, and building a system that watches them so a human does not have to.
Then I took an enterprise search and strategy role at a big agency, and for the first time I had genuine scale to work against. Enough data, enough surface area, enough recurring manual process that the builder reflex I had been running on my whole life suddenly had somewhere to point.
That was where the instinct I had been running on since I was eleven finally met a problem big enough to break it wide open.
Sundays
Here is the specific moment it all clicked, as best as I can reconstruct it.
I was sitting at the kitchen table on a Sunday in the fall of 2022, a couple of years before the twin girls showed up, and I had a spreadsheet open that I was dreading touching on Monday morning. It was a weekly client report. It was maybe my fiftieth time putting that same report together. The data lived in four different tools, the narrative was the same every week with different numbers, and I was going to spend ninety minutes on it Monday the way I had spent ninety minutes on it every Monday for the past year.
Something finally snapped. I closed the spreadsheet, opened a code editor, and wrote my first real Python script against an API.
Not to be dramatic about it, but that weekend was the before-and-after line for me. Within a month I had automated three of my most painful weekly tasks. Within a quarter I was fielding "can you build me one for my team" requests from people I did not even work with directly. Within six months I realized the actual highest-leverage thing I could do with my week was not the job on my business card anymore, it was the tools I was building in the margins.
Nobody handed me the title "marketing engineer." The title showed up in the industry maybe eighteen months later. When it did, I recognized it the way you recognize your own handwriting.
Credit where it is due on the naming: the team at Profound are the ones who put a proper stake in the ground and formalized "Marketing Engineer" as its own category, including running a whole Marketing Engineering curriculum to codify the skills and mindset. They did the hard work of giving the role a name, a definition, and a community. My story here is about what the work looked like before it had a label, and what it looks like from a practitioner who was doing it in the margins of other jobs for years. If you want the formal definition of the role, read theirs. What you are reading here is the first-person version.
When the Company Noticed
The first Sunday-to-Monday smuggling operation was a weekly report. Then it was a brief-drafting tool. Then it was a small research pipeline I had quietly running in the background. Eventually the tools I had been building on the side became the tools my day-job teams were running every day.
Credit where it is due here: Terakeet gave me room to run with it. At some point my employer turned the whole arc into a four-part career spotlight on LinkedIn, "The Spark," "The Innovation," "The Support," "The Balance." The honest read of that story is that most companies would have told me to stay in my lane. Terakeet did not, which is most of the reason any of the tools got built at all.
Over time the work pulled other people in. DMs asking "can I ask you a quick AI question." Prototypes landing in product-team meetings to show what had become possible in the last six months. Ad-hoc sessions on "can you teach my team how to prompt properly," pings about "we are hitting a wall with this workflow, can you look at it," asks like "I built this thing but it is messy, can you help me productionize it." The role started shaping itself around internal demand, not a job description.
That transition, from building for my own week to building for teams I did not even sit next to, is the real inflection point of the role. The first phase is self-serving: you build the tool that gives your own Friday back. The second phase is when other people start using it, then rebuilding around it, then asking for more, and an internal platform exists that did not a year ago. That only keeps going when the company formalizes it — when there is structural support for what had been a side-of-desk habit. That kind of support is what turns a curious builder into a functioning marketing engineer, and it is the step most companies do not realize they need to provide.
The Thing AI Actually Changed
Here is the part of the story I think about the most.
I was a builder my entire life. I had ideas in notebooks going back to when I was twelve. The reason most of them never became real is that real software was genuinely hard and expensive. To build the tool I wanted for the social-media network, I would have needed to hire a developer I could not afford. To build the ordering-error detector I had sketched for a construction project a decade ago, I would have needed a year and a data team. To build the glossary-of-every-weird-term tool I kept wanting, I would have needed a backend, a frontend, a deployment, a CMS.
AI did not make me more of a builder. AI removed the one thing that had been standing between me and every idea I had been carrying around since I was a kid: the friction of actually shipping the software.
Now I can sit down on a Wednesday night, after the girls are asleep, and have a working prototype of an idea I had a decade ago on my laptop by eleven. Not a slide about it. Not a Notion page about it. A thing that runs.
That is the actual revolution, and it is not really about "AI automation" or "AI agents" or whatever the vendor-speak is this month. It is that people who have been wired to build things their entire lives can now build things at the rate their brains move, instead of the rate software engineering historically moved at. The bottleneck was never the ideas. The bottleneck was the cost of turning them into reality.
How It Shows Up in the Work Now
The stuff I build these days is the adult version of the neighborhood handyman jobs, and I mean that in the most respectful way possible.
Someone has a need, I see the specific shape of it, I build a system that closes the loop. The cleanest public example is an AI ordering system I helped ship for a 50-person commercial door installer that now catches errors before they cost the company roughly $400K a year.
It also looks like the house I live in. I automated the litter box, the lights, the fans, the vacuum, because the reflex does not turn off when I clock out. It looks like the rare-coin scanner I built for a jar of old change, the coffee app that dials in a pour based on the specific beans and scale, and the walled-garden internet I put together for my twin daughters so they could explore without me worrying about what was a click away. Longer write-ups live at /builds.
All of those are the same instinct running on the same loop: see a thing that annoys me, build the thing that fixes it, ship it by the end of the weekend.
Why I Think This Matters for Whoever Is Reading This
A decent number of the people who land on this page are running marketing or growth teams and wondering what to do about the word "AI." My honest answer, having lived through this specific shift, is that you are not looking for a technology to adopt. You are looking for the person on your team who already has the reflex I have been describing, and who has been waiting on someone to give them permission to use it.
They are on your org chart right now. They built the little tool everyone on Slack asks them to re-share. They have a side project they will casually mention if you ask. They are probably underleveraged in their current role because the role was written before any of this was possible. If you give them a day a week, a small budget, and explicit permission to build, you will watch them change the shape of your team's output inside a quarter.
The tools have finally caught up to the instinct. If you have been sitting on ideas in a notebook for years waiting for the technology to be cheap enough, the waiting is over. You can ship one this weekend.