Machine Learning Engineer | Remote

Crossing Hurdles is seeking a PhD Rater for a part-time position focused on designing and evaluating real-world STEM benchmark problems. The role involves implementing tasks in Python, analyzing AI model behavior, and improving benchmark quality for AI models.

Responsibilities:

• Design challenging, real-world STEM benchmark problems in domains such as data science, machine learning, finance, and software engineering
• Implement tasks within an agentic development environment using Python
• Create reproducible problem setups with clear specifications and executable tests
• Evaluate and analyze AI model behavior, including reasoning traces and agent workflows
• Diagnose reasoning failures, logic gaps, and problem-solving limitations in AI systems
• Contribute to improving benchmark quality and evaluation frameworks for frontier AI models

Requirements:

• Active or recently graduated PhD
• Deep expertise in data science, machine learning, finance, and/or Python-based software development
• Strong research background in advanced STEM topics
• Ability to commit reliably for 30+ hours per week
• Demonstrated technical output such as high-quality open-source contributions or research work
• Ability to analyze agent behavior traces and diagnose failures beyond surface-level errors