Arabic evidence intake
Paste transcripts, upload audio, or record live sessions with consent-aware Arabic transcription.
Governed AI for enterprise assessment centers
Arabic assessment evidence, transformed into defensible English decisions.
Maqyas helps assessors capture Arabic sessions, run a governed competency pipeline, resolve human-review gates, and release evidence-backed reports with a complete audit trail.
No score without evidence
Human resolution audit trail
Owner, org, shared, and admin access
QC locked release workflow
System overview
Built around the assessor journey, not a black-box score.
From evidence capture to final release, Maqyas keeps the assessor in control. AI drafts the analysis, but gates, edits, staleness, and release decisions remain explicit.
Governed 9-stage pipeline
Every output carries status, confidence, traceability, and review gates from AC-P01 through AC-P09.
English reports with evidence
Final reports include Arabic verbatim evidence, English translation, decision node, QC state, and disclosure.
Operating model
One workspace for capture, review, governance, and release.
The system supports the full assessment center lifecycle: assessor preparation, session evidence collection, AI-assisted competency interpretation, human QC, and report delivery.
Assessor workspace
Draft assessments, resume incomplete evidence packages, and continue live pipeline runs from one dashboard.
Enterprise administration
Manage organizations, active users, assessor roles, and controlled sharing across teams.
Progress intelligence
Track completion, recent runs, draft freshness, gate status, and export readiness at portfolio level.
Measurement discipline
Keep competency ratings linked to source evidence, confidence, rationale, and reviewer decisions.
Governance model
PASS, REVIEW_REQUIRED, BLOCK: visible at every stage.
PASS
Evidence is sufficient and defensible.
REVIEW_REQUIRED
A human reviewer must resolve uncertainty.
BLOCK
Processing or release is halted.
Assessment workflow
Designed for repeated assessor operations.
Maqyas is structured for real teams running many candidates, not a one-off prompt. Each phase preserves context so teams can pause, inspect, and continue without losing accountability.
Create a candidate assessment with role, purpose, exercises, consent, and source materials.
Capture Arabic transcript evidence through paste, upload, or live recording.
Run the governed AI pipeline and inspect every stage output before release.
Resolve review gates with rationale, then export an English report with Arabic evidence.
Candidate report
QC passArabic evidence lens
تعاملت مع الأولويات بهدوء ووضعت خطة واضحة للفريق.
“Handled priorities calmly and set a clear plan for the team.”
Report release
Every recommendation ships with evidence, rationale, and model disclosure.
Maqyas produces English reports that remain connected to the original Arabic evidence. Assessors can review the decision node, competency ratings, confidence, QC notes, and export-ready output.
Access Maqyas