What the work actually involves
On a live project you receive batches of samples — raw text, source documents, or model-generated answers — plus an annotation guideline that defines the label set and the edge cases. Your job is to apply the guideline consistently and to say why. Most tasks ask for a label plus a short written justification, and many ask for a citation to an authoritative source in your field. Where a sample is ambiguous, mislabeled at the source, or outside the guideline's scope, you flag and escalate rather than guessing — reviewers treat a well-written escalation as a signal of quality, not a failure to complete work.
Because AfterQuery recruits across every professional domain rather than for one vertical, the assessment is generalist in structure but expects you to demonstrate depth somewhere specific. The strongest applicants name a domain, show real hands-on years in it, and can defend a judgment call under follow-up questioning.
What the screen is looking for
- Verifiable expertise, not adjacency. Years of practice, the kind of decisions you were accountable for, credentials where the field has them.
- Guideline discipline. Whether you can follow a rubric you disagree with, and whether you know the difference between an error and a legitimate ambiguity.
- Written reasoning. Annotations are read by other people; terse, specific, well-cited justifications carry more weight than long ones.
- Realistic availability. Batch work arrives irregularly and often has turnaround windows.
Logistics
Fully remote and asynchronous. The qualifying assessment is unpaid and is the gate to the pool — passing it does not guarantee an immediate project, only eligibility for matching. Paid work is project-based with flexible hours; observed rates span $25–60/hr, varying by domain scarcity and task complexity, and are not guaranteed. Expect no fixed weekly minimum, but also no steady pipeline — treat it as supplemental work that arrives in waves.