Leads the team this role joins
Hires for this team
On the team, can refer you
From your own Gmail and LinkedIn.
About the role
This is a fully remote contract role for machine learning engineers who have already done paid work training, evaluating or benchmarking AI systems. The work is building demanding technical tasks for advanced AI models, grading what those models produce, and setting the standard for a correct solution. The board lists the position under the United States, and it asks for at least 15 hours a week on a flexible schedule.
What you’ll do
- Design demanding machine learning engineering tasks that test advanced AI models.
- Write task instructions, reproducible environments, reference solutions and grading criteria.
- Assess model-generated code, experiments, analyses and other technical output.
- Spot data leakage, flawed validation, unstable training, performance problems and reproducibility gaps.
What they’re looking for
- Machine learning engineering background with prior paid work in AI training, evaluation, benchmarking or improvement.
- Working proficiency in Python.
- Hands-on use of PyTorch, JAX, TensorFlow, scikit-learn, XGBoost, LightGBM or comparable tools.
- Public evidence of skill, such as GitHub projects, competition results, Hugging Face work, publications or open-source ML contributions.
- Availability to begin immediately.
Nice to have
- Kaggle Expert, Master or Grandmaster status, or strong results in another technical competition.
- Experience writing AI training tasks, rubrics, graders or benchmark problems.
- Experience reviewing other AI trainers' technical submissions.
- Experience with long-horizon or agentic AI evaluations.
What’s on offer
- Contractor engagement, fully remote.
- Flexible schedule with at least 15 hours per week.
- Output-based pay: experts are paid for tasks that meet specifications and quality requirements, and task volume and completion time can vary.
Skills
Questions about this role
Is this role remote?
Yes, the posting describes it as fully remote.
How many hours per week does it need?
At least 15 hours per week on a flexible schedule.
How is the work paid?
Pay is output-based, earned by completing tasks that meet project specifications and quality requirements.
Do I need prior AI training experience?
Yes, the posting says it is intended for candidates who have already completed substantive AI training work.
What is the hiring process?
Submit your application, AI training experience and public technical evidence, complete an AI interview with technical and integrity checks, then a hiring manager review, followed by onboarding if selected.