Notes from the operator side.
Long-form writing on AI-powered job search, Claude Code skill design, and what the data from 740+ listings actually says. Less marketing, more methodology.
What 740 listings taught me about the AI-era job market
I evaluated 740 listings, applied to 68, and accepted one offer (Head of Applied AI at Zinkee). Here is what the numbers say about archetype fit, comp band realities, the tailoring response delta, and the patterns that bite candidates without an explicit rubric.
The complete guide to AI-powered job search in 2026
A practical 2026 guide to AI-powered job search. Covers the asymmetry between recruiter-side and candidate-side AI, the four phases of a structured search, which tools fit which user profile, and where to spend time vs where not to.
Why career-ops
I built career-ops to manage my own job search after exiting a 16-year operator role. 740 listings evaluated, one Head of AI offer landed. Then I open-sourced it. Here is what the project is, what it deliberately is not, and why this asymmetry needed a tool on the candidate side.