Experience:
7+ Years of Relevant Scientific Experience, or 3+ Years with at Least One Scopus-Indexed Publication

Job Type
Contractor

Location:
Remote Required Profiles: LinkedIn profile required to apply Role Overview We are looking for experienced scientific researchers and computational science professionals to develop and validate terminal-based tasks used to evaluate AI agents on scientific workflows. You’ll contribute your expertise in scientific computing, numerical modelling, simulation, and research validation to ensure tasks are technically sound and reproducible. Significant: Prior hands-on experience working on Terminal-Bench 3.0 projects is mandatory. A PhD is required; candidates currently pursuing a PhD are eligible.

Key Responsibilities
• Design multi-step terminal tasks based on realistic scientific and research workflows.
• Build self-contained computational environments with pinned dependencies and scientific software.
• Prepare experimental datasets, simulation parameters, and scientific model configurations.
• Implement reference solutions and automated tests for numerical and scientific correctness.
• Develop tasks involving simulation, numerical modelling, data fitting, optimization, spectroscopy, molecular analysis, or scientific visualization, as relevant to the specialization.
• Define numerical tolerances, units, boundary conditions, and expected scientific behavior.
• Debug dependency conflicts, solver instability, precision issues,



and file-format problems.
• Validate reproducibility and document research assumptions, limitations, and technical decisions. Key Requirements
• Bachelor's degree from a university ranked among the top 100 institutions in the country.
• PhD in a relevant scientific discipline; candidates currently pursuing a PhD are eligible.
• Prior hands-on experience with Terminal-Bench 3.0 projects.
• Relevant background in Physics, Chemistry, Biology, Mathematics, Statistics, Neuroscience, Medicine, Materials Science, or a related scientific field.
• Meet either of the following experience criteria:
• 7+ years of relevant professional or research experience; or
• 3+ years of relevant experience and at least one publication in a Scopus-indexed journal.
• Practical expertise in scientific or computational workflows.
• Comfortable discussing research methods and technical validation in an interview.
• Availability to work remotely on a flexible schedule. Preferred Technical Skills
• Scientific programming and computational modelling.
• Numerical simulation, optimization, and data fitting.
• Scientific visualization and domain-specific software.
• Automated testing and numerical validation.
• Research Software Engineering, scientific benchmarks, or reproducible research packages. Selection Process Step 1 – Short Screener Step 2 – AI Interview Step 3 – Paid Work Trial Step 4 – Project Onboarding

📌 Science Experts | $300/task (India)
🏢 Codefeast
📍 India