What the work actually involves

You build the test, not take it. A typical task starts with an authentic artifact you construct or de-identify — a line-item budget, a call sheet, a vendor agreement, a delivery schedule, a crew roster — and a scenario layered on top: a location falls through two days before principal photography, a post house misses a QC deadline, an overage appears in a cost report that doesn't reconcile with the hot costs. You then write the correct decision path, spell out why competing options are wrong, and decompose your reasoning into a rubric of 35 or more discrete criteria a grader (human or model) can apply consistently.

The rubric is where most of the effort lives. "Producer would push the day" is not a criterion; "identifies that pushing the day triggers a second holdover fee on the equipment package quoted in Exhibit B" is. Expect iteration with the project team on task design, calibration on rubric granularity, and revisions when a task turns out to be ambiguous or solvable by guessing.

What the platform screens for

  • Credited experience. Five-plus years with named projects and your actual function on them — not adjacency to production.
  • Hands-on financial and logistical ownership. Whether you personally reconciled budgets, issued call sheets, and negotiated or administered vendor and crew contracts.
  • Constraint reasoning under follow-up. Screeners push on your scenarios: what changes if the union is different, if the budget tier drops, if the deliverable is a live broadcast rather than episodic.
  • Written precision. Documentation must survive technical review by people who don't know production.
  • Realistic throughput. Minimum weekly submissions apply; the screen probes whether your stated hours are plausible against your other commitments.

Logistics

Fully remote, contractor, asynchronous — no fixed shifts, but real deadlines and a weekly submission floor. Roles are typically filled within about 48 hours, and selected experts are expected to submit first tasks within 24–48 hours of onboarding. Compensation is per accepted task, so effective hourly rate rises as you internalize the rubric format; the $45–85/hr band reflects what contributors have reported, not a guarantee. No prior AI experience is expected or required.