Job description
- $60–$120 per hour equivalent, paid per accepted task
- 100% remote
- Part-time
- Under 20 hours a week, you choose when
Turn competition-grade data science into paid, remote AI training work, on your own schedule.
AI coding agents are now asked to solve real data science problems, and someone has to decide whether their answers actually hold up. That takes people who have done the work: who know how a leaderboard solution is built, where leakage hides and what a reliable test looks like. In this role you design difficult tasks, write the reference answer, and check how AI agents perform against it. Your reviews help improve how AI coding agents handle quantitative work and code.
What you'll do
- Package data science tasks with their data, dependencies and run steps so anyone can reproduce them
- Write and verify reference solutions that show how each task should be done
- Design tests and grading criteria that catch incomplete or flawed answers
- Run AI coding agents on your tasks, study where they fail, including data leakage and wrong assumptions, and refine the task
- Why this is worth your time
- Earn $60–$120 an hour equivalent while designing data science tasks and checking how AI coding agents solve them
- Work on AI coding agents directly, testing their data science solutions and seeing exactly where they break
- Build practical experience turning competition methods into reproducible benchmarks, tests and clear verification notes
- Fully remote and flexible, no fixed shifts, no commute
Requirements & qualifications
- Who this is for
- Hands-on data science, machine learning, quantitative modeling, or research competition experience
- A competition history on Kaggle, DrivenData, Zindi, AIcrowd, Numerai, DataSource.ai, CodaLab/Codabench or a similar platform
- Strong Python: you can test, debug and read code you did not write, and implement and check solutions on your own
- Strong problem solving, with a clear grasp of why the methods in your solutions work
- Professional fluency in English
- Helpful background: published competition solutions, notable rankings, applied math, ML engineering or quantitative research.
- No experience in AI training work needed. Your competition experience helps test and improve how AI coding agents solve data science problems.
- Pay and details
- $60–$120 per hour equivalent; pay is per accepted task, as an independent contractor
- Expected commitment of about 15 hours a week, under 20; you choose the hours and days
- Remote, open to candidates in 52 countries, listed on the job page
Benefits
- Flexible hours
- Remote work
How to apply
- Apply on OpenTrain (upload your resume): https://app.opentrain.ai/api/out/cmv6wfi0r000c06plu0j9do5…
- Complete a short application on the hiring platform.
- If shortlisted, take a conversational AI interview with technical questions and a coding challenge, whenever it suits you.
Ready to put your competition experience to work on AI task design and evaluation? Apply on OpenTrain.
About OpenTrain AI
OpenTrain AI connects experienced professionals with paid, part-time AI training work: shaping and improving the AI systems that companies are building right now.
PAID OPTIONS: NONE.
Salary & compensation
About OpenTrain AI
OpenTrain AI connects experienced professionals with paid, part-time AI training work, helping shape and improve the AI systems that companies are building right now.