Verified Remote Opportunities

Description

Full Job Description

• Location: Remote

• Type: Contract / Part-time

• Commitment: 10-20 hours per week

• Compensation: Up to 40 USD / hr

• Project duration: 2 months, with potential extension

About the role

We create high-quality STEM training data for frontier AI models that is directly used in training and evaluation workflows at leading AI labs.

We are looking for experts in Chemistry to design advanced, deterministic problems with exactly one verifiable correct answer. These problems should challenge modern AI systems with deep reasoning, domain knowledge, and computational rigor.

Projects may include computational chemistry, physical chemistry, analytical chemistry, chemical engineering-adjacent workflows, and chemistry tasks requiring specialized software tools.

What you’ll do

• Design difficult chemistry problems that reflect real scientific workflows

• Create deterministic tasks with one correct answer and full verified solutions

• Develop reasoning-intensive and computationally grounded problems

• Use Python and, where appropriate, domain-specific chemistry tools or simulation packages

• Ensure all tasks are reproducible, technically precise, and clearly documented

• Write concise and rigorous technical explanations in English

What we’re looking for

• Master’s or PhD in Chemistry or a related field

• Research or industry experience involving computational or quantitative chemistry workflows

• Strong Python skills; familiarity with scientific libraries

• Strong understanding of modeling, numerical approaches, and chemical reasoning

• Ability to design original, difficult problems that go beyond textbook recall

• Excellent attention to detail and clear technical writing

Nice to have

• Experience with computational chemistry, cheminformatics, spectroscopy analysis, molecular modeling, or specialized chemistry software

• Experience with professional chemistry workflows involving simulation or data analysis tools

• Experience creating technical evaluation content