AI Trainer (ML Engineering) and Task Designer

You will turn practical machine learning experience into technical tasks that test advanced AI systems. The work includes designing tasks, preparing reproducible environments and reference solutions, and judging whether model-generated technical work is correct and reliable.

• Create realistic machine learning engineering tasks for advanced AI systems.
• Build reproducible environments, reference solutions, evaluation criteria, rubrics, and graders.
• Review model-generated Python code, experiments, analyses, and other technical deliverables.
• Identify data leakage, incorrect validation, unstable training, performance problems, and reproducibility issues.
• Assess technical submissions against clear project specifications and quality requirements.

This is a remote contractor role for an expert-level contributor with strong practical machine learning experience and the ability to set objective standards for correct technical solutions.

• $60 to $120 per hour equivalent, paid per task that meets the project specifications.
• Flexible part-time schedule, with a commitment of at least 15 hours and a listed requirement of less than 20 hours per week.
• Fluent in English.
• Prior paid experience training, evaluating, benchmarking, or improving AI systems.
• Strong practical Python skills.
• Hands-on experience with PyTorch, JAX, TensorFlow, scikit-learn, XGBoost, LightGBM, or comparable machine learning tools.
• Public technical evidence through repositories, model implementations, competition results, publications, technical reports, benchmarks, or open-source contributions.
• Helpful experience includes creating AI-training tasks, reference answers, rubrics, graders, or benchmark problems; reviewing technical submissions; technical competitions; machine learning research or open source; and evaluating long-horizon or agentic AI systems.