Productized offer · Hiring intelligence
Shortlist reads every CV against what the role actually needs, then hands you a ranked six — each with the evidence that earned the rank and the question to probe in the interview. Every rejection carries a reason you'd say out loud. Blind by default.
Blind pass complete — same six, same scores. The rubric never saw a name.
The problem
You posted a role on Friday. By Wednesday there are 214 applications and a hiring manager with a day job. What happens next isn't screening — it's triage under fatigue.
Strong candidates don't announce themselves. The career-changer with exactly the right project history sits at position 163 — and nobody reads to 163.
Screening happens after the real work, when you're tired. Fatigue makes conservative calls: familiar employers get through, unusual careers get binned. That's when bad decisions get made.
"Didn't feel right" is not a reason — it's a liability. A rejection you can't articulate is a legal and moral mess, and it quietly teaches your team to hire people who look like themselves.
How it works
What does this role actually need? We build a scored rubric from your past successful hires and your definition of good — weighted, written down, signed off by you. No rubric runs without your name on it.
We run Shortlist on a role you already filled. If it wouldn't have surfaced your best hire from that original pile, we say so — and you walk away having paid nothing.
Every new application is read against the rubric within an hour of arriving. The ranked shortlist updates daily; every "no" carries its reason. You spend your time on six conversations, not 214 documents.
The difference
Same applicant pool, two methods. Left is what "5+ years Python" plus gut feel produces. Right is a rubric calibrated to your definition of good.
The filter keeps the keyword-stuffer who wrote "Python" eleven times and shipped nothing. It bins the career-changer who'd have been your best hire because her CV says "Ruby" in 2019.
Then gut feel finishes the job — at 11pm, in your own image. Rejection reason on file: none.
Every candidate is scored on evidence against your rubric — years where they matter, scale where it matters, gaps flagged as interview probes rather than silent vetoes.
Blind by default: names, photos, and addresses never reach the scorer. Each of the 214 outcomes — including every "no" — comes with a reason you could read to the candidate's face.
Pricing
No per-seat pricing, no percentage of salary, no incentive to inflate the shortlist. You know the number before a single CV is read.
Questions
Both, by design. Screening runs blind: names, photos, ages and addresses are stripped before scoring, and the rubric only contains criteria you signed off as job-relevant. Every rejection carries a specific, evidence-based reason — the audit trail most hiring processes wish they had. And the human decides: Shortlist ranks, you choose. Nothing about a candidate's fate is fully automated.
Knowledge work with 30+ applicants per posting — engineering, data, product, design, marketing, operations, finance. If the role has a definable "good," it can be rubriced. If the role gets six applications total, you don't need me; read them over coffee.
No — the ad stays yours. But calibration usually surfaces two or three lines in your posting that attract the wrong pile, and I'll say which. Most clients rewrite the ad, watch the next pile improve, and count that as a free extra.
CVs are processed on infrastructure I control and document, under a data-processing clause written in plain English. Nothing is used to train anyone else's model, nothing is retained past the engagement, and deletion is part of offboarding. If you have compliance requirements, we design around them during calibration week.
Good — that's the system working. Every score decomposes into evidence you can inspect line by line, so disagreement is a conversation about criteria, not a fight about vibes. If a criterion was wrong, we fix the rubric and rescore the pile in a day. You always make the final call; the rubric just has to earn it.
Fit
Next step
Tell me what you're hiring for and roughly how big the pile usually gets. I'll tell you honestly whether Shortlist pays for itself on this role — and if it doesn't, what to do instead.
Replies within one business day — rubric talk, no sales deck.