Learning to Build, Not Just Hire For It
An argument by Rik Avalos
For most of my career, my job was to find the person who could build the thing, not to build it myself. Tesla, Neuralink, Teladoc: in every one of those rooms, I was the person evaluating builders, not one of them.
That changed this year.
The gap I kept hiring around
Every company I've worked with had the same shape of problem: a founder or exec with a clear idea and no easy way to see it built without first raising money, writing a job description, and waiting months for the right person to say yes. I spent two decades closing that gap for other people.
I never expected to close it for myself.
What actually changed
The honest answer is Bob Michaels, my co-founder at Trinzik, decided that mentoring me in AI development was worth his time. Not as a favor, as a bet that an operator who understands people, positioning, and go-to-market could get genuinely useful at building if the tooling got out of the way.
Claude Code turned out to be that tooling. This site, the Rik Avalos Band site, and a growing list of small projects exist because I sat down and built them, with an AI collaborator standing in for the engineering team I would have spent months hiring for.
I'm not pretending to be an engineer. I'm pretending, less and less, to need one for everything.
Why I'm writing this down
I'm starting to publish essays like this one alongside short conversations, the audio version of this piece is linked below, because the most useful thing I can offer people watching AI change how companies get built is a real account of doing it, not a polished one.