What the work actually involves

You write problems that a competent AV tech could solve on a job site and that a language model probably cannot. That means scenarios with real constraints: a 48kHz Dante network dropping clock sync mid-rehearsal, a three-camera corporate shoot where the switcher output is a frame behind the program audio, a hypercardioid placement fighting a ceiling reflection in an untreated room. Each item needs a defensible correct answer, a written rationale that survives argument, and usually a set of plausible-but-wrong paths that show why they fail.

The second half of the job is evaluation. You read model output on other people's problems and score it — flagging where the answer is technically wrong, where it is right for the wrong reason, and where it recites a best practice that would not survive contact with the actual gear. Written feedback is expected to cite the specific standard, spec, or physical principle at stake, not just say "incorrect."

What the screen looks for

  • Hands-on specificity. Screeners push on details: gain structure numbers, connector pinouts, latency budgets, color space handoffs. Vague familiarity surfaces fast under follow-up.
  • Breadth across environments. Live event, studio, broadcast, corporate AV and post each have different failure modes. Candidates who have only ever worked one are still viable but are asked to be honest about the boundary.
  • Writing under constraint. Can you state a troubleshooting sequence in a paragraph without hedging, and can you say what would make your own answer wrong?
  • Grading judgment. Distinguishing a wrong answer from an unconventional but workable one is the core skill; over-penalizing valid alternate approaches is a common disqualifier.

Logistics

Fully remote and asynchronous — there is no shift and no fixed call time. Contributors typically commit a defined block of hours per week and are paid hourly for accepted work. No studio access or equipment is required; the work is written, though a working knowledge of current tooling is assumed. Expect calibration rounds early on where your scoring is compared against other experts before volume increases.