Not theory — the exact loop and scoring we run today, with the failure modes we’ve hit named and addressed.
Four loop stages; each with named signals and counter-signals.
Backend + frontend. Each ~4 hours. Solutions included.
What to ask, what to follow up on, what to flag.
How to stop hiring against the loudest interviewer.
Notion + PDF. Use as-is, fork the format, or just lift the parts that improve your current loop.
60+ page PDF: the loop, the scoring, the calibration, the rejection reasoning. Reads start-to-finish in a couple hours.
Per-candidate scorecard, loop tracker, calibration meeting agenda. Drop into your workspace and start using.
Backend (Go) and frontend (React). Includes reference solutions and per-criterion scoring.
Three real (anonymized) scorecards: a strong yes, a strong no, and a borderline hire — with calibration notes.
If something isn’t answered here, ask in your intro email — we keep this list short on purpose.
Partially. The loop structure transfers; the scoring is calibrated for senior. We include a section on adjusting weights for mid-level — junior is its own thing we don’t cover.
The take-home prompts in the pack are distinct from what we use in our own loop today — so candidates can’t google a leaked version. If you want fully unique prompts, we offer a custom variant (email us).
The pack covers software-engineering hiring broadly. AI-specific evaluation (eval design, prompt judgment, model debugging) isn’t covered — that’s a different rubric we haven’t productized yet.
Across our team: 12+ year average tenure, 1-in-80 offer rate, sub-5% regrettable attrition. Yours will differ — the rubric is necessary, not sufficient.
We do bespoke kits, too. Send a paragraph about the problem and we’ll come back inside 48 hours.