Contract Review AI Evaluator — Remote Contractor

AuraOne; United States; On-site; USD 27,552 - 55,104; Full time

Job summary

AuraOne in the United States is seeking an experienced AI data reviewer to turn contract evaluation outputs into auditable labels, rationales, and regression cases for AuraOne Human Data. This contractor role focuses on calibrating model results against a versioned rubric and tagging issues such as hallucinations and policy violations.

You will compare paired responses, document recurring failure modes, and maintain reviewer-quality scores weekly.

Qualifications

• Experience with evaluation or annotation of contract review AI systems.
• Comfort applying multi-page rubrics consistently across long batches.
• Clear written reasoning that names the issue and the rubric clause 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

• Evaluate contract review AI evaluation model outputs against a versioned rubric and assign severity tags for Contract Review AI Evaluator assignments.
• Compare paired responses and pick the stronger answer with a written rationale.
• Label hallucinations, instruction-following failures, and unsafe content with structured tags.
• Capture ambiguous prompts and route them back to the program team for rubric updates.
• Maintain reviewer-quality scores by calibrating against gold-standard examples each week.
• Document recurring failure modes so the modeling team can target them in the next training run.

Skills

Model output evaluation

Rubric annotation

Severity tagging

Inter-rater calibration

Contract evaluation

Legal reasoning

Policy review

Risk analysis

Contract review

Review

Job description

AuraOne in the United States is seeking an experienced AI data reviewer to turn contract evaluation outputs into auditable labels, rationales, and regression cases for AuraOne Human Data. This contractor role focuses on calibrating model results against a versioned rubric and tagging issues such as hallucinations and policy violations.

You will compare paired responses, document recurring failure modes, and maintain reviewer-quality scores weekly.