Enterprise TA · Selection science · ATS governance

High-volume hiring. Evidence intact.

A governed 4,000-applicant architecture connecting automated knockout controls, structured assessment, recruiter capacity, fairness monitoring, and privacy without turning scale into a black box.

100% synthetic · demographic fields isolated from every hiring decision

Applicants

4,000

2,400 Reference · 1,600 Focal

Conversion

3.0%

120 accepted hires / 4,000 applied

Time to fill

28.5d

Mean across 5 requisition families

Feedback SLA

91.8%

1,836 / 2,000 at or below 48 hours

Adverse impact ratio

0.87

Knockout progression · passes 4/5ths

Halo-effect control · CAND-2026-0013

The impression was stronger than the governed evidence.

The +0.68 gap triggers review without diagnosing intent. The candidate was not hired because evidence did not support an override.

Subjective4.60

Composite3.92

Gap+0.68

Enterprise decision system

Counts, rates, and stage definitions stay visible together across five requisition families.

Recruitment funnel & stage drop-off

4,000 applied → 986 shortlisted → 500 assessed → 120 hired.

Structured evidence vs subjective intuition

All 500 finalists; amber identifies a gap at or above +0.50.

Evaluation turnaround & SLA by department

Average feedback hours with percent submitted at or below 48 hours.

Adverse impact by stage

Cohort selection rates with lower-to-higher impact ratio.

High-performance candidate pipeline

Search and filter all 4,000 records; only the current page enters the DOM.

Candidate	Requisition	Furthest stage	Composite	Bias gap	Outcome	Evidence

1–25 of 4,000 candidates