Generalists

About this role

Mercor is hiring highly detail-oriented Generalists on behalf of an AI lab to support **human data pipelines used to train and evaluate advanced AI systems**. This role is ideal for people who are extremely quality-focused, methodical, and take pride in producing consistently high-accuracy work. You’ll work on structured evaluation, labelling, and review tasks where precision matters more than speed. Small mistakes compound here — we’re looking for people who notice edge cases, inconsistencies, and ambiguity instinctively. ### **What You’ll Do** - Review, evaluate, and annotate AI outputs with extreme attention to detail - Follow complex guidelines precisely and apply them consistently across tasks - Identify subtle errors, edge cases, and quality issues others might miss - Provide clear, structured feedback to improve downstream AI training quality - Maintain high accuracy across repetitive, high-volume workflows - Flag unclear instructions or ambiguous cases proactively ### **What We’re Looking For** - Exceptionally **detail-oriented** and quality-obsessed mindset - Prior experience as a **CB / reviewer / rater** in human data or AI training pipelines - Proven ability to follow nuanced instructions without deviation - Strong written communication and reasoning skills - High reliability, consistency, and comfort with repetitive precision work - Ability to work independently and maintain focus over long task batches ### **Nice to Have** - Experience with large-scale AI data annotation or evaluation programs - Background in QA, operations, trust & safety, or structured review work - Familiarity with multi-step rubric-based evaluations - Comfort giving calibrated feedback aligned with strict quality bars ### **Why Join** - Work directly on training next-generation AI systems - Clear expectations and quality-first culture - Long-term opportunity for consistent, high-quality contributors

Eligible applicant countries

This role accepts applicants from:

• USA

Skills & domains

• ai-training
• rlhf
• sme
• annotation
• Miscellaneous