AI Recruitment & Talent Acquisition

Overview

Screening at volume is where hiring quality and fairness are usually lost. This module screens and ranks candidates against the criteria in the approved job description and shows the reasoning behind every fit score: why this candidate, always. Recruiters keep the decision; the system keeps the evidence.

What it does

• Criteria-based screening

  Candidates screened against the approved job description: industry experience, years, qualifications, and skills scored dimension by dimension.

• Explainable fit scores

  Every score carries its reasoning. Why this candidate, always shown, to the recruiter and in the record.

• Pipeline tracking

  Shortlists, interviews, and offer states held in one pipeline, so no candidate advances or stalls silently.

• Impact-ratio analytics

  Selection rates monitored across groups using the impact-ratio math defined by NYC Local Law 144.

How a run works

Every module runs the same contract: the system drafts, a named human approves, and the approved artifact keeps its history.

1. 1Draft

   The module screens and ranks the applicant pool into a shortlist; identifying details are pseudonymized into neutral tokens before any external model call.

2. 2Human gate

   A recruiter reviews the ranked shortlist with the reasoning behind each score shown; advancing, rejecting, or extending an offer is a human decision the system cannot take alone.

3. 3Approved artifact

   Approved shortlist and offer decisions are recorded with their reasoning, and accepted offers hand off to lifecycle and payroll with the trail intact.

Proof

• WHY THIS CANDIDATE

  Fit scores are explainable by construction. The reasoning is stored with the score and always shown.

• LL144 IMPACT RATIO

  Bias and impact-ratio analytics use the same math New York City Local Law 144 audits require.

• PSEUDONYMIZED

  Candidate identities are replaced with neutral tokens before any external model sees the data.