Astrophysics & Cosmology Expert

$70–100/hrrate as listed

Design graduate-level computational astrophysics and cosmology problems that test whether frontier AI models can run real research workflows in astropy and related tools. Each task goes through a testing loop against state-of-the-art models, and you iterate on it until it lands at the target difficulty.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

The work

You are writing original problems, not labeling existing data. A typical task starts from a research workflow you have actually run — a cosmological distance or growth calculation, an angular power spectrum estimate, a galaxy survey selection-function analysis, a reduction pipeline step — and turns it into a fully specified problem with a verifiable answer. You write the setup, the oracle function that produces the ground truth, and the validator that decides whether a model's answer is correct. A second, harder class of task hides information: the model must plan a sequence of queries or simulated measurements to recover something not directly visible, which forces strategic reasoning about what to measure and how to narrow the space.

After the problem is built, you run it against current frontier models. If they solve it immediately, it is too easy; if they fail for uninteresting reasons — ambiguous wording, a broken environment, a numerically unstable oracle — that is your problem to fix. Expect to iterate, and expect reviewers to push back on tasks whose difficulty comes from tedium rather than reasoning.

What the screen looks for

Verifiable hands-on tool use. Publications, repos, or professional work showing you have written real code against astropy, CAMB/CLASS, healpy, emcee, or comparable libraries — including knowledge of their edge cases and failure modes.
Python engineering competence. There is a coding assessment; you will be writing oracles and validators, not notebooks you alone can read.
Problem-design instinct. Can you articulate why one version of a question is hard for good reasons and another is merely long?
Comfort in Linux/terminal and remote compute sandboxes, including reproducible environments.

Logistics

Fully remote and asynchronous, with at least 15–20 hours per week expected and sustained availability valued over bursts. Observed pay for this listing is $70–100/hr, typically set by assessed depth and domain scarcity; rates are as reported and not guaranteed. Graduate training (MS or PhD, or equivalent research experience) is effectively a floor for this cohort.

What they look for

Graduate-level training in astrophysics, cosmology, or a closely related field (MS/PhD or equivalent research experience)
Hands-on research coding with astropy and related tools (CAMB/CLASS, healpy, emcee, survey reduction pipelines)
Strong Python — writing problem setups, oracle functions, and automated solution validators
Comfort in Linux/terminal environments and remote compute sandboxes
Demonstrable evidence of tool proficiency via publications, open-source contributions, or professional work
Benchmark, exam, or problem-set design experience (nice to have)

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you hold a Master's degree or PhD in a STEM field, or do you have equivalent research experience (e.g., multiple years as a research scientist or engineer)?
Graduate-level STEM training · required
Have you written and run code using astropy, or similar astrophysics libraries, as part of research, professional work, or published open-source contributions?
Hands-on astrophysics software proficiency · required
Do you have hands-on experience writing Python code in a research or professional context (not just coursework or tutorials)?
Python coding experience · required
Are you able to work in a Linux command-line environment and use remote compute resources or sandboxes?
Linux/terminal competency · preferred
Are you available to commit at least 15–20 hours per week to this project?
Availability · required

0 of 5 answered

Included with this listing

Interview prep — 11 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether you've actually written production code in this domain, not just run tutorials—can you articulate the workflow, the libraries' behavior, and the judgment calls you made.
A strong answer: Name a specific problem (cosmological distance calculations, survey data reduction, power spectrum analysis), describe the setup and constraints, explain which astropy functions you chose and why alternatives wouldn't work, and show you understand what the code actually outputs and how you validated it.
Where people lose it: Describing only the science or only the programming in isolation; vague references to 'using astropy' without showing you've debugged or adapted it; inability to explain why you picked one approach over another.