Skip to main content
Candidate Screening • Rubric Intelligence

Score every candidate on the same objective rubric.

Paste your JD, and RecruitAI drafts a 10-point multi-dimensional rubric automatically. Every candidate is graded blind against identical criteria — with matched skills, confidence scores, gaps and tailored interview questions rendered side-by-side.

From JD to calibrated rubric — in seconds

JD Analysis

Extracts role title, required skills, seniority and competencies. No manual rubric authoring.

Weighted Dimensions

Five pillars — Core Skills, Architecture, Problem Solving, Communication, Velocity — adjustable with ±5% steppers.

Consistent Calibration

Every candidate sees identical questions and weights, eliminating ad-hoc interviewer variance.

Side-by-side screening matrix

Stop debating gut feelings in spreadsheets. RecruitAI’s matrix maps candidates on the same scale with green ≥80% / amber 50-79% / rose <50% fit tiers. At a glance you see verified strengths, gaps and suggested interview probes for each applicant, exportable to PDF for hiring committees.

What screening surfaces
  • Stack mastery with source snippet citations
  • Years of depth per competency, with timeline validation
  • Impact metrics — scale, throughput, users, revenue affected
  • Flagged concerns: gaps, short tenures, scope mismatches
What it replaces
  • 15–20 hours per requisition of manual skimming
  • Inconsistent phone screens and unstructured feedback
  • Bias from name, photo or university prestige
  • Copy-pasted scorecards with no evidence

FAQ — Candidate screening

What makes this different from generic ATS scoring?▾

RecruitAI generates a rubric from your actual job description — not a fixed template. Scoring dimensions (technical mastery, system design, communication, velocity) are calibrated to your role and weighted by importance, with every score backed by extracted evidence from resumes.

Can we customize rubrics per role?▾

Yes. Edit weights (Core Skills, Architecture, Problem Solving, Communication, Velocity) via ±5% steppers, toggle blind mode per campaign, and reveal PII only when you move a candidate to interview.

How do you prevent hallucinated experience?▾

All agent outputs are validated against strict Pydantic schemas, and skill matches include source snippets from resume chunks. Candidates also receive red-flag detection for timeline gaps and unverifiable claims.

Is scoring auditable for compliance?▾

Every score, blind redaction decision and status change is timestamped and stored with tenant isolation. Bias audit endpoint reports selection-rate impact ratios for NYC LL144-style reviews.

Score your next 100 candidates free

Bring your JD and resumes. Get a calibrated rubric and blind scores in minutes — no talk with sales needed.

Create Free Account