What the work actually is
You receive batches of model outputs responding to prompts about CRISPR/Cas systems, gRNA design, delivery, off-target assessment, and adjacent gene-editing topics. For each one you make two separate judgments that people often collapse into one: is this scientifically correct, and does it carry dual-use risk? A response can be technically wrong and harmless, or accurate and inappropriate to surface. You annotate against a rubric, flag biosecurity concerns, and write a short rationale that a reviewer who is not a bench biologist can follow and audit. The written justification is the deliverable as much as the label — thin rationales are the most common reason work gets sent back.
What the screen is looking for
AfterQuery's screening is AI-led and leans on verifiable specifics. Expect to be asked what you personally did at the bench rather than what your lab does: which Cas variants, which design tools, whether you ran GUIDE-seq or CIRCLE-seq yourself or read someone else's output. Follow-ups drill into the reasoning behind choices — why that PAM, why that validation method, what your off-target hit rate looked like. Familiarity with DURC frameworks, biosafety policy, or the gene-drive and enhanced-pathogen literature is preferred rather than required, but candidates who can reason about where the line sits between textbook molecular biology and operationally sensitive detail tend to score well.
Logistics
- Fully remote and asynchronous; no standing meetings, no fixed hours.
- One-week project window, 10–20 hours inside it, work assigned project by project.
- Compensation is observed at a flat $1,000–$1,200 per week-long project, with up to $1,500 reported for exceptional candidates — a band, not a guarantee.
- No long-term obligation; repeat projects depend on annotation quality on the first batch.
This suits a Year 2+ PhD student or a strong Masters researcher with real hands-on time in a gene-editing lab. It does not suit someone whose CRISPR exposure is coursework or a single rotation.