Machine Learning Research Expert

This role involves creating research-grade machine learning problems that probe the real capabilities of frontier AI models, along with the solutions and scoring criteria used to judge them. It suits published ML researchers holding or pursuing a Master's or PhD who write model code as part of their regular work. The role is with an AI company that accepts experts into skill pools, which project teams then draw from as work comes up.

What you'll do

• Build realistic ML research problems within your own specialty
• Write expert solutions and rubrics that set the standard for grading model answers
• Work in areas such as language models, reinforcement learning, robotics, ML systems, time series, causal reasoning or AI for science

What you'll need

• A completed or in-progress Master's or PhD in machine learning, computer science, statistics, maths or another quantitative discipline
• At least a year of hands-on ML research and one or more first-author ML papers
• Practical experience with training, evaluation or systems code; labelling work does not count
• Preferred: several first-author papers, measured gains over baselines, or publications at venues like NeurIPS, ICML, ICLR or ACL

Good to know

• Fully remote and asynchronous, with 5–40 flexible hours per week
• Listed pay is $150 per hour
• Background checks are part of the review, and complete applications are reviewed first