Hi. I do two things:
I write software and I help
engineers use AI well.
I've spent my career building production software — these days mostly in Elixir and Phoenix, but I'm comfortable across Node, Ruby, TypeScript and Python. I take on senior full-stack contracts for teams that need another pair of experienced hands.
Alongside that, I teach AI — not to non-technical audiences, but to other software professionals. Workshops, conference talks, and working sessions about what current models actually do well, how to fold them into engineering workflows, and what to be skeptical of.
Senior engineering work
Full-stack Elixir & Phoenix, with regular forays into Node, Ruby, TypeScript, and Python. Greenfield product builds, backend systems, and rescue work on codebases that have outgrown their original shape. Contract or fractional.
AI talks & workshops
For conferences, all-hands, and engineering team off-sites. I talk about the current state of LLMs for software development — what works, what doesn't, and how to update your mental model without buying into the hype.
AI working sessions
Half-day to multi-day sessions with engineering teams, sitting down together with real code and real workflows. Less "here's a talk", more "let's figure out where this actually fits for you."
A handful of things I'm consistent about — for the engineering work and the AI work both.
-
Built, not slidewared. The AI work is run by someone who still writes production code daily. Examples come from real systems, not demos.
-
Model-agnostic. Claude, ChatGPT, Gemini, local — the principles transfer. I'll be specific about where they don't.
-
Honest about the limits. AI is a tool. I'll tell you where it helps your team and where it'll slow you down.
-
Short engagements. I'd rather earn the next call than lock in a retainer. Most engagements are a workshop or a 4–8 week build.
Working on something interesting? Tell me about it.
Email's the best way. I read everything; usually reply within a day or two.
bjorn@zeiler.solutions