Legal Expert for AI Training & Evaluation (Project-Based / Contract)

Weekday is an AI-native recruiting platform backed by Y Combinator and General Catalyst. We have built the largest white-collar talent database in India and have developed tools for outbound recruiting campaigns.

We are currently running an AI training and evaluation project as a YC-backed AI data lab, seeking legal experts to design and evaluate tasks for advanced AI models in legal work—research, contract review, and compliance.

Advanced AI models are being integrated into legal work, but their effectiveness depends on the quality of the tasks and evaluations provided by experts like you. You will:

• Create complex legal tasks—such as research questions, contracts for analysis, or compliance scenarios—with clear, defensible model answers
• Review AI-generated legal responses line by line, grading them on accuracy, reasoning, and completeness
• Identify hallucinated case law, misread clauses, missed exceptions, and flawed reasoning
• Write clear rationales for evaluations to help the model and team learn from mistakes
• Work across legal research, contract analysis, regulatory compliance, and legal reasoning—not limited to a single niche

Every task is reviewed, and you will be measured on the quality and rigor of your submissions, not on volume.

Requirements

What you'll actually do:

• Design complex legal problems with clear, defensible model answers
• Review AI-generated legal responses for accuracy, reasoning, and completeness
• Catch errors like hallucinated case law, misread clauses, and flawed reasoning
• Write precise rationales for evaluations
• Work across legal research, contract analysis, compliance, and legal reasoning

What we're looking for:

• 3+ years of relevant legal experience—practice, in-house, or at a firm
• Strong legal research and analytical skills to trace questions to the right authority and explain applicability
• Comfort with contracts, compliance, and reasoning across multiple areas of law
• Precision in writing—clear, concise, and free of padding
• Low tolerance for plausible-sounding errors in AI outputs or human reasoning
• Reliability in meeting deadlines for task-based work
• Curiosity about how AI will transform legal work and a preference for shaping its evolution

What you get:

• ₹20,000–25,000 per task, paid for depth of thinking rather than hours logged
• Remote, flexible work that fits around a practice or full-time role
• A front-row view of how frontier AI models are trained and evaluated on legal work, with the opportunity to improve them
• Access to a YC-backed team that operates quickly and maintains direct communication
• More tasks and larger projects if your work is consistently strong