Competitive Data Scientist

About the Role

micro1 is recruiting data scientists, machine learning engineers, and applied mathematicians with a background in competitive data science (Kaggle, DrivenData, Zindi, AIcrowd, Numerai, and similar platforms) to help build and evaluate challenging tasks for AI coding agents. This is a flexible, part-time contractor engagement of roughly 15 hours per week, with the schedule set by the expert.

What You'll Do

• Build reproducible task environments, prepare data and dependencies, and implement verified reference solutions
• Design evaluation criteria and tests that reliably separate correct solutions from flawed ones
• Run coding agents against tasks, analyze their failures, and refine the tasks accordingly
• Spot implementation errors, flawed assumptions, and data leakage that affect correctness
• Document task requirements, reasoning, and verification steps clearly

Requirements

• Demonstrated experience in data science, ML, quantitative modeling, or competitive platforms such as Kaggle or DrivenData
• Strong problem-solving skills and a clear understanding of your own methods
• Practical programming ability, ideally in Python, including testing and debugging unfamiliar code

Compensation & Logistics

Pay is $60-$120/hour, fully remote, roughly 15 hours/week with a flexible schedule (weekends optional). Note: the screening questions indicate actual pay may be output-based (per completed task meeting spec, with a weekly minimum task quota) rather than strictly hourly — confirm exact terms before accepting.