george larson

Who is George

The thread isn't the title.

It's not about the job title. It's not about gaming a keyword filter.

It's about craftsmanship.

Titles came and went. The thread is this: take the messy work, the kind that's expensive when it breaks, automate the ripe parts, dig into the harder problem. I've done this. On the record. A manufacturing business had customers designing heat transfers with an AI tool I shipped in 2023, alongside an internal knowledge platform and a bridge across two vendor pipelines.

Not a pitch, actual history. Honest then, honest now: if the work can be automated, automating it is the work.

A vibecoder ships the happy path and demos the glossy bits.
The interesting part is the failure not advertised in demos.
The provider that stalls mid-call without throwing an error.
The task runner wedged on a channel that never times out.
The reconciliation that quietly disagrees with itself.
Most of the serious work is found in those cracks, in the parts that don't screenshot well.

Evidence over adjectives.

The case studies and the open-source record are the projects. Each one clicks through to the real thing.

Successful record running the tech and the teams both.
When a downturn eliminated the infrastructure-director role, I absorbed that team and its full stack, no added headcount.
Engineers and non-engineers alike.
You can't lead where you don't go.

AI: used heavily, and stated plainly.
The proof it's used well is that it's not obvious.
AI lets a careless person ship faster, but a careful one go deeper.
That difference is found in the work, not a banner.

Hiring, or stuck on something? Let's talk.