AI Evaluation Reviewer - Administrative Ops (Remote)

AuraOne Human Data; United States; Remote; USD 28,000 - 41,000; Part time

Job summary

AuraOne, Inc. is seeking an Administrative Assistant (AI Evaluation) to work remotely from the United States.

The role focuses on evaluating AI outputs within assistant specialist operations workflows, ensuring policy adherence, and judging workflow quality. You will grade tone, escalation logic, and stakeholder fit using a structured rubric, flag risks, and document actionable next steps for model training.

Qualifications

• Direct working experience in assistant specialist operations on real teams for Administrative Assistant (AI Evaluation) work.
• Comfort applying multi-page rubrics consistently across long batches.
• Clear written reasoning that names the policy or workflow being applied.
• Strong attention to detail and the ability to flag when a prompt itself is the problem.
• Reliable async availability for at least 10 hours per week.

Responsibilities

• Review AI outputs against current assistant specialist operations workflows, playbooks, and firm policy for Administrative Assistant (AI Evaluation) assignments.
• Grade tone, escalation logic, and stakeholder fit on a structured rubric.
• Flag operational risk, missed escalations, and policy-adherence gaps with severity tags.
• Capture the right next step so the modeling team can train on it.
• Adjudicate disputed treatments against published playbooks or firm guidance.
• Maintain reviewer-quality scores in inter-rater calibration cycles.

Skills

Operational review

Policy adherence

Workflow judgment

Stakeholder communication

Assistant specialist operations

Job description

AuraOne, Inc. is seeking an Administrative Assistant (AI Evaluation) to work remotely from the United States.

The role focuses on evaluating AI outputs within assistant specialist operations workflows, ensuring policy adherence, and judging workflow quality. You will grade tone, escalation logic, and stakeholder fit using a structured rubric, flag risks, and document actionable next steps for model training.