There's a particular kind of tired that comes from reading your 14th resume of the evening and realizing you already forgot what the first five said. When you're hiring solo, there's no recruiter in the middle softening the edges — just you, a job description you wrote, and a pile of PDFs that all open with "results-driven professional."
CandidateScore does one thing and shows its work. You paste a resume and the job description. It returns a structured fit scorecard: a fit score on a 1-10 scale, the candidate's strengths relative to the JD, the gaps it found, and — this is the part I cared about — the evidence behind every single call. Not "strong culture fit," but the line in the resume that earned the point.
The evidence requirement is the whole point. A score with no receipts is just a vibe with a number attached. When the tool flags a gap, it has to show you the requirement it measured against; when it credits a strength, it points at the matching experience. You still make the call. The tool just makes sure you're comparing every candidate against the same written bar instead of against your mood at 11pm.
What it deliberately does not do
- No ranking guarantee. A 1-10 fit score is a structured comparison aid, not a prediction of who will succeed on the job. Two strong candidates can both score high; the score doesn't decide for you.
- No claim that it reads intent. It matches the resume to the JD you provided. If your job description is vague or contradictory, the scorecard reflects that — it won't invent qualifications the candidate didn't state.
- No stored candidate database. You paste; it scores; nothing is kept. If you need a pipeline, that's a different tool.
The limitation I'd want to know as a user: the score is only as good as the job description you feed it. A one-line JD gives a one-line-quality scorecard. Spend the effort on the posting and the tool returns that effort multiplied across every applicant.
CandidateScore is $19/mo. Paste a resume and a JD and read the scorecard before your next phone screen — and trust the evidence column more than the number.