About the Role
Help evaluate realistic US-law tasks and documents used to train and assess large language models. This expert-review role focuses on improving the accuracy, realism, and quality of legal evaluation data, rather than client-facing legal work.
What You'll Do
- Review US-law tasks and documents for legal accuracy, completeness, and realism.
- Provide structured feedback on flawed samples and recommend precise corrections.
- Rework legal examples to meet quality standards.
- Draft practical guidelines to improve future legal evaluation datasets.
What We're Looking For
- At least five years of substantive professional experience in a legal field.
- Legal training and professional practice in the United States, with demonstrated knowledge of US law.
- Experience reviewing legal documents and providing clear, precise written feedback.
- Experience consulting on AI systems, evaluating AI-generated legal content, or creating legal training or evaluation data is advantageous.
Compensation & Benefits
Compensation is $80 to $110 per hour, depending on relevant experience and seniority.
Location
Fully remote. US legal training and practice are important; current residence is flexible.
