What the work actually involves

You write realistic office management problems and then judge how well a model handles them. A task might start as a messy scenario — a two-person front desk with overlapping PTO requests during an audit week, a vendor invoice that doesn't match the signed service agreement, a hybrid seating plan that breaks when headcount grows 20% — and your job is to specify the correct handling, the defensible alternatives, and the answers that sound plausible but would cause real problems. The second half of the work is evaluation: reading model responses and writing feedback that says precisely where the reasoning failed, not just that it did.

Good items in this category are hard because office management is full of decisions with competing constraints: budget versus coverage, policy consistency versus individual circumstance, speed versus documentation. Content that only tests definitions ("what is a purchase order?") gets rejected. Content that forces the model to sequence actions, weigh trade-offs, and account for who needs to be informed is what gets used.

What the platform screens for

  • Evidence you have actually run an office — headcount supported, systems owned, budget scope, vendors managed, whether you supervised staff directly.
  • The ability to explain why a decision was right, in writing, with the constraints named. Screeners follow up on vague answers.
  • Calibration in grading: can you distinguish a response that is wrong from one that is merely different from how you would have done it.
  • Range across settings. Someone who has managed a 12-person law office, a clinic front office, and a 90-person startup writes more varied items than a single-employer specialist.

Logistics

Fully remote and asynchronous — no standing meetings, work claimed and completed on your own schedule. Contributors commonly work in blocks of a few hours; consistent weekly availability matters more than volume, because reviewers calibrate against your prior submissions. Pay is hourly within the observed $50–120/hr band, set by project and demonstrated experience, and is not guaranteed at any point in that range. Expect a written sample task as part of onboarding.