What the work actually involves

This is not writing prompts or grading chatbot outputs. You are paired with another professional and you coach — a real session, spoken aloud, recorded or transcribed for use as training data. The scenario might be a mid-career analyst weighing a pivot into product, a laid-off manager rebuilding a search after eight months, or someone rehearsing answers for a panel interview next week. You run the conversation the way you would with a paying client: open with discovery, ask questions before you advise, reflect back what you heard, and land on something the person can act on.

Sessions typically run 30–60 minutes. You may be asked to cover a specific scenario type from a project brief, stay inside a defined coaching modality, or avoid certain content (compensation negotiation specifics, legal advice on terminations). Guidelines also tend to police realism — filler, hedging, and the messiness of a genuine session are usually wanted, not scrubbed out. Expect to re-record if a session drifts off-scenario or one party breaks character.

What the screen is looking for

  • Actual coaching practice, not adjacent HR or recruiting work. Screeners probe for caseload, session structure, and how you handle a client who resists the obvious answer.
  • Questioning discipline. The single most common failure is a coach who advises in the first two minutes. Reviewers listen for whether you diagnose before prescribing.
  • Range across scenarios — early career, mid-career transition, involuntary exit, internal advancement — and the judgment to change register between them.
  • Guideline compliance. Can you hold a scenario constraint for 45 minutes without narrating that you are following a scenario.

Logistics

Contractor, remote, United States. Paired sessions are live, so this is scheduled rather than fully async — you pick availability blocks and get matched with a partner. Volume is project-driven and lumpy: a heavy stretch followed by weeks of nothing is normal. You need a quiet room, a decent headset, and reliable bandwidth, since audio quality directly affects whether a session is usable. Pay is stated at $50–100/hr as observed on the platform, not guaranteed, and typically varies with scenario complexity and how much depth your sessions demonstrate.