One person. One AI. 740 job offers evaluated.

What this is

A person built a personal system where an AI — specifically Claude — handles the exhausting parts of job hunting. It reads job offers, scores them on ten different factors (like salary, growth potential, culture fit), spits out a grade from A to F, and then writes a custom CV tailored to that specific role. Not a generic CV. One shaped around what that company is looking for.

The system can process ten or more offers at the same time, running in parallel the way a small team would.

Why it's interesting

It's not a product. No monthly fee, no onboarding call. Someone built this for themselves, stress-tested it across 740 real job listings, and ended up landing a senior role at the end of it. The thing actually worked.

What's quietly remarkable is how it's configured: you don't have to dig into technical settings. You just tell the AI in plain language — "focus on marketing leadership roles instead" — and it rewrites its own instructions accordingly.

What it means for you

If you've ever posted a job and been buried in identical CVs, this shows you what's coming. And if you're hiring or being hired, the arms race between AI-written applications and AI-screening systems is already here.

Worth sitting with: what parts of your own work could benefit from this kind of structured, opinionated evaluation?

Words worth knowing

AI agent — an AI that doesn't just answer questions, but takes actions: reading documents, making decisions, producing outputs, one step after another.

ATS — Applicant Tracking System. The software companies use to filter CVs before a human ever sees them. Most large companies use one.

Sub-agents — smaller AI helpers that work on separate tasks at the same time, like having a few people each reading a different document in parallel.

Repo — short for repository. A publicly shared folder on GitHub where someone posts their project so others can see or use it.