Earth Sciences AI Benchmark Designer

OpenTrain AI, Inc.; United States; Remote; USD 47,000 - 87,000; Part time

Job summary

OpenTrain seeks an Earth Sciences AI Task Designer to convert authentic Earth-science workflows into reproducible terminal benchmark tasks and datasets. You will design environments with Python, R, Bash, or Julia, enabling rigorous evaluation of AI systems.

You will validate data formats, coordinates, and tolerances while documenting provenance and limitations. Remote contractor role with five-week duration and flexible hours.

Qualifications

• Advanced training in Earth Sciences or closely related field (doctorate/postdoc or equivalent).
• Strong programming skills and experience with scientific data processing and modeling.
• Proficiency in Linux/terminal environments and scientific workflows.

Responsibilities

• Translate Earth-science workflows into self-contained terminal benchmark tasks.
• Prepare scientific datasets and reproducible environments with appropriate libraries and tools.
• Create expert solutions using Python, R, Bash, Julia or related software.
• Define objective grading criteria, validate coordinates, units, timestamps, and numerical tolerances.
• Document data provenance, edge cases, and limitations.

Skills

Earth Sciences

Scientific programming

Python

Linux

Geospatial analysis

Education

Doctoral/postdoctoral in Earth Sciences

Tools

Python

R

Bash

Julia

Job description

OpenTrain seeks an Earth Sciences AI Task Designer to convert authentic Earth-science workflows into reproducible terminal benchmark tasks and datasets. You will design environments with Python, R, Bash, or Julia, enabling rigorous evaluation of AI systems.

You will validate data formats, coordinates, and tolerances while documenting provenance and limitations. Remote contractor role with five-week duration and flexible hours.