Legal AI Training & Evaluation Specialist

MPower Plus; United States; Presencial; USD 80.000 - 120.000; Jornada completa

Descripción de la vacante

MPower Plus is seeking a Legal Domain Expert to support AI training and evaluation projects in the United States. The role requires hands-on experience with prompts, evaluation rubrics, and assessing AI-generated legal content for accuracy and reasoning quality.

The ideal candidate will develop legally grounded prompts, create structured rubrics, and provide actionable feedback to improve AI models. This position emphasizes direct involvement in AI training and data quality review.

Formación

• Hands-on AI training, grading, or model-evaluation experience.
• Practical experience writing prompts and evaluation rubrics.
• Experience assessing AI-generated legal content for accuracy and reasoning quality.

Responsabilidades

• Develop legally grounded prompts and scenarios to test AI reasoning and responses.
• Write structured evaluation rubrics and grading criteria for legal accuracy, reasoning, completeness, and instruction adherence.
• Evaluate and grade AI-generated legal responses against established criteria.
• Identify inaccurate legal interpretations, unsupported conclusions, fabricated citations, and other model errors.
• Provide clear, actionable feedback and reference-quality responses to support AI model improvement.
• Apply legal subject-matter expertise to training-data development and quality review.

Conocimientos

AI training

Rubrics writing

Prompt development

Legal accuracy evaluation

Descripción del empleo

MPower Plus is seeking a Legal Domain Expert to support AI training and evaluation projects in the United States. The role requires hands-on experience with prompts, evaluation rubrics, and assessing AI-generated legal content for accuracy and reasoning quality.

The ideal candidate will develop legally grounded prompts, create structured rubrics, and provide actionable feedback to improve AI models. This position emphasizes direct involvement in AI training and data quality review.