Sovereign candidate vetting — your rules, on your servers

Stop pasting CVs into US-cloud AI. Assess candidates against your own written rules, on sovereign New Zealand infrastructure, with a guard against prohibited-grounds bias and a logged, human-decided audit trail.

Your vetting rules

Edit, add, or remove. The bias-guard and human-decision rules are locked — they are the legal backbone.

Assess a candidate

SOVEREIGN AI AUDIT ENTRY
candidate:
model:
ruleset:
residency:
decision:

What pasting a CV into US-cloud AI does

Candidate PII (name, DOB, contact, salary) crosses the border to a US company · no ruleset · opaque ranking · no audit trail · inputs may be retained for training · automated-hiring bias unguarded.

What Village does

Assessment runs on your sovereign NZ Situated Language Model · against your written rules · prohibited-grounds bias-guarded · every step logged with provenance · a human makes and records the decision · candidate data never leaves sovereign ground.

Demonstration — the assessment shown is illustrative. In a production Village it runs against your tenant's own Situated Language Model (villageai-14B on the sovereign NZ GPU) and writes a real audit-log entry.