ML Research Expert, Published First-Author Benchmark Author

Design realistic machine learning research tasks that test what AI models can do. Work in one or more areas such as language models, deep learning, reinforcement learning, computer vision, robotics, efficient ML systems, optimization, forecasting, causal reasoning, trustworthy learning, or AI for science.

• Write technically sound reference solutions for benchmark tasks.
• Create grading rubrics that distinguish strong model results from incomplete or incorrect work.
• Check that each task measures the capabilities it is intended to test, using your research judgment.

What it pays and takes

• Pay: $150 per hour.
• Schedule: Flexible, asynchronous contract work for 5 to 40 hours per week.
• Location: Fully remote; open to candidates in eligible countries only, not worldwide.
• Language: Fluency in English.
• Education: A completed or in-progress Master's or PhD in machine learning, computer science, statistics, mathematics, or another related quantitative field.
• Research: At least one first-author research paper in machine learning or a closely related field, plus at least one year of hands-on machine learning research experience.
• Technical skills: Ability to write machine learning code for training, evaluation, or systems work, with depth in at least one relevant research area.
• You must be able to explain technical decisions and judge results against meaningful standards. Multiple first-author publications or papers showing measured improvements over a baseline are helpful.

How it works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI training work

AI training work uses human-written examples and reviews to help build and evaluate AI systems. Researchers are paid for bringing the subject knowledge and judgment needed to create rigorous tasks and assess model results.