I'm not a lifelong programmer. That's the point.
Homegrown AI Consulting exists because of one uncomfortable, wonderful discovery: a person who knows their work deeply — and learns to ask AI precisely — can build real tools that solve real problems. I know, because that person was me.
Where it started
A few years ago I knew next to nothing about coding. What I had instead was the same thing you probably have: a pile of repetitive work, a suspicion there was a better way, and a growing annoyance that every "AI solution" I saw was either a sales pitch or science fiction.
So I started small. Not with a computer-science course — with one real problem, and a question: can I get AI to help me build a thing that fixes this?
The first lesson: sometimes the answer isn't AI
Here's a thing nobody selling AI will tell you. Some of my most valuable early builds were plain Python scripts that don't call AI at all. Once a task is truly repetitive — same steps, same rules, every time — a deterministic script does it perfectly, instantly, and for free, forever. Calling an AI model for that job means paying a little "token tax" every single run, for a result that's less reliable.
The AI's job was writing that script with me. My job was knowing the work well enough to describe it precisely, and knowing what "correct" looks like. The AI cooks; you fish. That division of labor is the entire secret, and it's learnable.
Then the builds got bigger
With the confidence from those first wins, I set up AI-assisted workflows for the recurring tasks that genuinely benefit from AI — the messy, judgment-adjacent stuff where a script alone isn't enough.
Then I took on something that would have sounded impossible to me a year earlier: over a gigabyte of raw public auto-safety data from NHTSA — the federal agency that collects vehicle complaints and recalls. Using Python text extraction, I surfaced part-type failure patterns that aren't explicitly recorded anywhere in the dataset, and rendered the results into clean HTML views a person can actually use. The full story is on the Proof page.
From there I went on to build brand-new tools for the automotive aftermarket — tools that simply didn't exist before.
Why I do this for other people now
Because the gap I crossed is crossable. It doesn't take a degree. It takes someone who knows the work, a bit of honest guidance about what AI can and can't do, and the one skill I now teach as the centerpiece of everything: asking precisely.
Small businesses are drowning in exactly the kind of repetitive work this approach eliminates — and they're the least served by an industry that only wants to sell enterprise "transformation." Homegrown AI Consulting is my answer to that: practical help, grown in your house, owned by you.
Your business has a story like this waiting.
The first chapter is a twenty-minute conversation. No jargon, no pressure — just an honest look at your work.