N
Nukhba
Responsible AI

How our AI works

Nukhba's AI screens, matches, and interviews using your rubric. It doesn't decide — it recommends. Every recommendation is auditable and reversible.

Responsible AI at Nukhba
UAE
Data Residency Sovereign Cloud
PDPL
Privacy Aligned Framework
SSO
Emirates ID Azure AD
A.I.
Ethical Bilingual Recruitment
ISO
27001-Aligned Controls
What you get

Everything you need to run this end-to-end

  • Retrieval-augmented scoring against your JD rubric
  • No black-box decisions — every score has a rationale
  • Human review required at every consequential step
  • Model updates published in the customer changelog
Deep dive

Nukhba's AI screens, matches and interviews using retrieval-augmented scoring against your JD rubric. It is grounded, explainable, and always advisory — never authoritative.

Every score cites its evidence. Every override is logged. Every model update is announced in the customer changelog before it ships to production.

Responsible AI at Nukhba
How it works

Three steps, always in this order

1

Ground in the JD

The rubric is derived from the requisition you approved — not from a black-box embedding of past hires.

2

Score with evidence

Each criterion returns a score with the specific evidence that supported it, and the gaps that lowered it.

3

Route to a human

Consequential steps — reject, invite, offer — always require a human decision. The AI drafts, the recruiter decides.

FAQ

What buyers ask most

No. Every reject, invite, and offer is confirmed by a human.

Ready to modernise your hiring?

Book a 30-minute walkthrough with a UAE recruitment specialist.