Generalist Expert (AI Training & Data Annotation)

Mercor, a talent marketplace backed by Benchmark and General Catalyst, is seeking Generalist Experts to support a large-scale project with a leading AI research partner. Your goal is to create and refine a diverse dataset of human-aligned tasks for Supervised Fine-Tuning (SFT) of frontier AI models. This is a short-term, project-based opportunity ideal for professionals with strong writing and analytical skills.

Key Responsibilities

• Prompt Generation: Create realistic base prompts across a range of generalist topics.
• AI Evaluation: Review and revise AI-generated responses for clarity, accuracy, and alignment with human reasoning.
• Iterative Refinement: Develop prompt-response iterations using variations in length, style, and system prompts.
• Benchmarking: Contribute to benchmarking efforts by identifying failure points in model responses.
• Quality Control: Apply consistent quality standards across diverse content types (summaries, Q&A, synthesis).

Requirements

• Experience: 2+ years of experience in research, education, writing, editing, or adjacent analytical fields.
• Skills: Proven ability to communicate complex topics clearly and concisely with strong attention to detail.
• Autonomy: Comfort working independently with evolving task guidelines.
• Tech Savvy: Familiarity with LLMs or AI-assisted writing tools is a plus.
• Visa Status: Please note, we are unable to support H1-B or STEM OPT candidates at this time.

Work Details & Benefits

• Schedule: Fully remote and asynchronous—set your own schedule.
• Commitment: Expected 10–25 hours/week.
• Duration: Approximately ~2 weeks with potential for extension.
• Payment: Paid weekly via Stripe Connect.