Skip to content
Rubrily

Fit Score

The Fit Score: explainable AI candidate scoring

One number to rank every candidate by — built from evidence, weighted your way, and traceable down to every justification. This page explains exactly how Rubrily scores candidates, because scoring you can't audit is scoring you can't trust.

Join the waitlist

What is a Fit Score?

A Fit Score is a 0–100% score that blends a candidate's AI CV evaluation and AI interview results into a single ranking number. Each component is scored 0–10 per criterion against your weighted rubric, with a written justification, so every Fit Score can be traced back to specific evidence.

How AI candidate scoring works in Rubrily

  1. 01

    Every criterion gets a 0–10 score with a written justification.

    The AI evaluates the CV and the interview independently against your rubric and writes down why it gave each score, citing what it found.

  2. 02

    Your weight tiers shape the rollup.

    Each criterion carries a tier — Must Have, Very Important, Important, or Good to Have — so a weak Must-Have hurts far more than a weak nice-to-have. Each engine rolls up to a weighted 0–10.

  3. 03

    You control the CV-vs-interview blend.

    Per project, set each component's intensity; Rubrily normalizes to 100% total weight. A code-heavy role can weigh the interview higher; a portfolio role, the CV.

  4. 04

    The blend becomes the Fit Score.

    CV and interview scores combine into 0–100%, shown with its full component breakdown. Weighting is project-scoped — the same assessment can produce different Fit Scores in different roles, because fit depends on the role.

Fit Score configuration showing CV and interview components weighted 50/50 with a normalized donut chart.

How are the weights set?

You set weights twice: per criterion, via four tiers (Must Have → Good to Have) that control how much each criterion moves its component score; and per project, via intensity sliders that control how the CV and interview components blend. Rubrily normalizes everything to 100% automatically.

When the AI can't evaluate, it says so.

Why "no fabricated scores"?

Because a guessed number is worse than no number. When there isn’t enough signal — a missing CV section, an unanswered question — Rubrily returns “Cannot evaluate” for that criterion instead of inventing a score. A missing answer is treated as missing, not as zero, so candidates aren’t silently punished for gaps in data.

Everything downstream inherits this: rankings, reports, and shared links only ever contain scores the AI could justify.

Click any score and see its reasons.

Every Fit Score decomposes: gauge → component scores → per-criterion 0–10 rows → written justifications → the transcript and recording they came from. The report adds a narrative summary citing specific evidence, plus Strengths, Gaps, and Recommendations for follow-up interviews.

Is AI candidate scoring fair and auditable?

In Rubrily, every candidate for a role is scored against the same rubric with the same weights, which removes the biggest inconsistency in manual screening. Every score carries a written justification you can audit, the full transcript and recording are one click away, and the hiring decision always remains with your team.

92%

Fit Score

Candidate report rubric rows with per-criterion scores and justifications.

Frequently asked questions

What is a Fit Score?
A Fit Score is a single 0–100% score that blends a candidate's CV evaluation and AI-interview results, weighted however you choose. It's fully traceable: click into any Fit Score to see the component scores, per-criterion breakdown, and the justification behind every number.
Can I change the CV/interview weighting?
Yes, per project. Each scoring component has an intensity you control, and Rubrily normalizes the total to 100%. Because weighting is project-scoped, the same candidate assessment can produce different Fit Scores on different roles — fit is always relative to the role.
What happens when the AI can't evaluate something?
It returns "Cannot evaluate" for that criterion instead of guessing, and the gap is visible in the report. Unscored criteria are excluded from the weighted math rather than counted as zero, so missing data never quietly sinks a candidate.
Do recruiters see why a score was given?
Always. Every criterion score comes with a written justification citing what the AI found, and the overall score carries a narrative summary plus Strengths, Gaps, and Recommendations. If a justification doesn't hold up when you read it, you can see that too — that's the point.

See it with your own role.

Join the waitlist

Define your rubric once — every candidate gets scored against it.