Materials Engineer & Python Expert - Freelance AI Trainer

Role overview

This project-based opportunity pairs domain experts with AI training work for major technology clients. The focus is on crafting, testing, and refining AI model challenges rather than joining a permanent team. Contributors work on specialized computational material science tasks that can be objectively checked through code.

What you will do

You will create material science problems that can be evaluated automatically and that depend on a specialized scientific tool. Suitable tools may include ObsPy, instaseis, pyrocko, MITgcm, flopy/MODFLOW, or similar domain packages. Simple data-cleaning exercises on artificial toy datasets are not considered a fit.

• Select a core tool and build a problem that depends on its specialized kernels, inversion methods, flow solvers, or validated scientific workflows.
• Develop a Python reference solution and provide any required input data, model files, or domain definitions.
• Define the correct numerical result and the tolerance the model must meet for a solution to count as accurate.
• Run repeated batch tests against the model, adjusting difficulty until the success rate lands in the target range.
• Submit the task for senior review in your specialty area once it is calibrated and performing within the desired range.

Calibration and working style

Task tuning is expected to take patience and iteration. The goal is to reach a pass rate of roughly 10% to 30% through changes such as rewriting waveform setups, tightening inversion settings, or refining solver tolerances. This process helps build deeper familiarity with the selected tool and provides practical insight into how frontier AI systems approach seismic, oceanographic, and subsurface-flow problems.

Applicant profile

This role suits material scientists or engineers who already know Python and are looking for part-time, non-permanent project work. You should be comfortable building tasks that truly require a specialized solver and working independently if you need to learn a new scientific package.

Qualifications

• A degree in material science or a closely related field.
• At least 2 years of experience in research, applied work, or teaching.
• Strong Python skills for writing reference implementations.
• Working knowledge of, or a willingness to rapidly learn, at least one scriptable package such as ObsPy, instaseis, pyrocko, MITgcm, xmitgcm, flopy/MODFLOW, or GeoPandas.
• The ability to design problems that genuinely require a specialized solver rather than a generic approach.
• Excellent written English at C1 level or above.

Application instructions

Please send your CV in English and mention your English proficiency level in your application.

How the process works

The workflow is: apply, complete qualification step(s), join a project, finish assigned tasks, and receive payment.

Time commitment

During active project periods, the expected workload is around 10 to 20 hours per week. This is only an estimate and is not guaranteed. It applies only while a project is active.

Compensation

Contributors may earn up to the equivalent of $45 per hour, depending on experience and the pace of contribution. Pay can vary from one project to another based on scope, difficulty, and required expertise.

Additional information

Participation is project-based rather than permanent employment.