1
Criteria before candidates
Role requirements are documented before resumes are reviewed.
Reduce bias risk through workflow, not slogans.
The stronger standard is bias-resistant, human-led, auditable review — not a promise that software can eliminate bias.
Bias Control Layer
Slate avoids “bias-free” claims. It uses controls that reduce risk, improve consistency, and make reviews auditable.
2
Evidence-only outputs
AI summaries must point to resume evidence or missing verification items.
3
Risky language scanner
Unsupported feedback is flagged before it becomes part of the record.
4
Human decision gate
People decide whether to advance, hold, reject, or route for manual review.
Risky language scanner demo
This demo flags vague, subjective, or potentially discriminatory manager notes before they are saved.
Click “Scan note” to check the sample manager note.
Alternate review safeguard
Manual review path
Route candidates to human-only review when AI assistance is inappropriate.
Protected info guardrail
Do not use protected or non-job-related details to score candidates.
Documentation warning
Slate should store a clear record of who reviewed and why.