The work
You will turn real life sciences research methods into multi-step computational tasks used to evaluate AI agents. The work focuses on scientific reasoning and execution, not simple question answering.
Tasks run in controlled, network-isolated computational environments. You will create the task materials, build expert solutions, check scientific accuracy, and make the results objectively gradable.
- Design challenging computational workflows across the life sciences.
- Create realistic input files, scientific datasets, instructions, constraints, and expected deliverables.
- Build reproducible expert solutions, ground-truth answers, and automated or semi-automated grading criteria.
- Create tasks that require data inspection, method selection, code execution, troubleshooting, and synthesis of results.
- Validate scientific assumptions, calculations, code, intermediate outputs, and final answers.
- Keep tasks self-contained and executable in a provided sandbox.
- Maintain task quality and respond to reviewer feedback.
What it pays and takes
This is a remote contractor assignment focused on computational scientific analysis. The assignment lasts five weeks.
- $280 per approved task.
- Expected commitment of 20+ hours per week.
- At least six hours per day and four hours of overlap with Pacific Time.
- Ph.D., postdoctoral training, or equivalent research experience in the life sciences.
- Strong scientific programming experience, particularly in Python.
- Experience with multi-step computational scientific analyses and validating computational outputs.
- Ability to turn authentic research workflows into bounded, objectively gradable tasks.
- Familiarity with scientific libraries, command-line tools, or reproducible research pipelines.
- Helpful backgrounds include bioinformatics, computational genomics, systems biology, computational neuroscience, biostatistics, computational drug discovery, computational biochemistry, structural biology, protein engineering, and computational microbiology.
- Professional fluency in English.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training work
AI training work uses human expertise to prepare examples and evaluations that help artificial intelligence systems improve. Experienced researchers are needed for specialist projects because they can check whether code, methods, calculations, and scientific conclusions are sound.