Operations engineer, welder, market trader, LA SMMA owner, ed-tech founder. Then TrainYourAgent. Twelve months from zero to $20K MRR — the wins, the bad bets, the offer that finally worked, and what I'd do differently.
I get DMs every week from founders trying to build the same thing — a custom AI agent practice serving SMBs. The number ($20K MRR) sounds small to the SaaS-poisoned founder twitter crowd, and it sounds large to the freelancer crowd. The truth is it's a meaningful milestone for the kind of business it is: services, owned-IP, recurring, profitable from month two. Here's the honest story of how I got here. The detour that turned out to be the moat Before TrainYourAgent, I had five business chapters that all looked unrelated: Operations engineer at a multi-site industrial company — learned to instrument processes, write SOPs that actually got followed, and respect the difference between "looks automated" and "is automated." Welder — three years on the floor.
It was: customers will pay a premium for someone who shows up on time, finishes the job, and doesn't blame the materials. Market trader — futures, options, a bit of crypto when crypto was new. Learned to make decisions with incomplete information and to size positions when I was wrong. This is the single most useful skill I have as a founder. LA SMMA (social-media marketing agency) — three years running paid ads for service businesses in Southern California.
Operations engineer, welder, market trader, LA SMMA owner, ed-tech founder. Then TrainYourAgent. Twelve months from zero to $20K MRR — the wins, the bad bets, the offer that finally worked, and what I'd do differently. It is filed under Founder Notes because that is where operators looking for this problem actually start, and it is written from production work rather than from a content calendar.