Company
Technology
STEMComputationalScientificSoftware&EvaluationDesign-Astrophysics&Cosmology
Neural analysis suggests this role is
optimal for Mid+ candidates.
“STEM Computational Scientific Software & Evaluation Design - Astrophysics & Cosmology. Skills: Computational astrophysics, Cosmology, Scientific software, AI evaluation. Design computational astrophysics problems. Design cosmology problems”
Industry & Context.
Analytical thinking; Problem design intuition; Iterative experimentation
15–20 hours per week
What They're Looking For.
Must Have
Graduate-level education in STEM field, Master’s, PhD, or equivalent research experience, Hands-on scientific computing tools experience, Apply computational methods to research, Advanced Python programming skills, Deep understanding of astrophysical modeling, Deep understanding of cosmological calculations, Deep understanding of observational data analysis, Distinguish challenging scientific reasoning problems, Comfort working independently in Linux, Comfort working in terminal-based environments, Comfort working on remote compute systems, Analytical thinking, Problem design intuition, Iterative experimentation mindset, 15–20 hours per week availability
Nice to Have
Experience with astropy, Experience with astrophysical software libraries, Experience with cosmological software libraries, Publications, Research projects, Open-source contributions
What You'll Do.
Design computational astrophysics problems
Design cosmology problems
Develop evaluation tasks
Create multi-step reasoning challenges
Implement problem setups
Implement oracle functions
Implement validation logic
Refine problem setups
Refine oracle functions
Refine validation logic
Iterate on task difficulty
Calibrate with AI models
Analyze model failure modes
Improve evaluation task robustness
Improve evaluation task clarity
Improve evaluation task scientific rigor
Refine benchmark design standards
How You'll Work.
Team & Collaboration
Collaborate in refining benchmarks
Full Job Description
## Accountabilities Design advanced computational astrophysics and cosmology problems using real scientific workflows, including simulations, observational data pipelines, and cosmological calculations. Develop evaluation tasks that test AI systems’ ability to use tools such as astropy and related scientific computing libraries. Create multi-step reasoning challenges involving data interpretation, hidden-variable inference, and experimental strategy design. Implement and refine problem setups, oracle functions, and validation logic using Python. Iterate on task difficulty through calibration cycles with frontier AI models to ensure appropriate challenge levels. Analyze failure modes of models to improve robustness, clarity, and scientific rigor of evaluation tasks. Collaborate in refining benchmark design standards for scientific reasoning and computational problem-solving. Requirements: Graduate-level education (Master’s, PhD, or equivalent research experience) in Astrophysics, Cosmology, Physics, or a closely related STEM field. Strong hands-on experience using scientific computing tools such as astropy and related astrophysical or cosmological software libraries. Proven ability to apply computational methods to real research problems, ideally demonstrated through publications, research projects, or open-source contributions. Advanced Python programming skills for scientific computing, simulation workflows, and evaluation tooling. Deep understanding of astrophysical modeling, cosmological calculations, or observational data analysis techniques. Ability to distinguish genuinely challenging scientific reasoning problems from superficially complex ones. Comfort working independently in Linux/terminal-based environments and remote compute systems. Strong analytical thinking, problem design intuition, and iterative experimentation mindset. Availability of at least 15–20 hours per week for ongoing project work. Benefits: Competitive hourly compensation ranging from $70
Applying for this STEM Computational Scientific Software & Evaluation Design - Astrophysics & Cosmology role?
Most applicants get filtered before a human reads their resume. See if yours makes the cut.
How to Apply on Lever
- Lever uses a streamlined one-page form — apply in under 5 minutes.
- LinkedIn import works well; review parsed data before submitting.
- The cover letter field is optional but visible to reviewers — use it to differentiate.
- Referral codes from employees can significantly boost visibility of your application.
ANONYMOUS · UNFILTERED
What do employees actually say about this company?
Real rants from real employees. Read before you apply.