Clinical Mental Health Expert — AI Conversation Evaluation

$100/hrrate as listed

Mercor is recruiting licensed and clinically trained mental health professionals to judge which of two simulated therapy-style conversations reads as more authentic. The task itself is short — roughly 30 minutes, fully remote and asynchronous — with an observed rate of $100/hr and the possibility of being tapped for later clinical evaluation work.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work actually involves

You log into an online interface and are shown pairs of simulated clinical mental health conversations — exchanges between a clinician figure and a patient or client. Your job is a forced comparison: which of the two reads like something that could have happened in a real session, and which one betrays itself. That means noticing the things a clinician notices and a layperson doesn't — a client who discloses trauma history in paragraph two with no ambivalence, a therapist who reflects perfectly every time, risk language that never gets followed up, a session that resolves cleanly in twelve turns. The whole task is scoped at up to 30 minutes. It is a discrete evaluation batch, not an ongoing engagement, though Mercor uses these tasks to build a pool for future clinical AI research work.

What the screen is looking for

Mercor's application flow is AI-led: a structured interview plus credential and background checks against your uploaded CV. For this listing the gates are real clinical training and real clinical hours — the platform wants evidence you have sat in the room, not just read about it. Expect probing on what setting you practised in, what population, and how you would describe the texture of a difficult session. The second thing being measured is evaluation judgment: can you separate "this conversation is realistic" from "this conversation demonstrates good therapy"? Those are different questions and the task asks the first one. Candidates who collapse them tend to score poorly.

Logistics

Remote and asynchronous. No scheduled calls, no live sessions with the research team.
Short duration. One batch, capped around 30 minutes of evaluation time.
Independent work. No calibration partner, no consensus round — your rating stands alone, which is why the screen weights independent qualitative judgment.
Rate. $100/hr is the observed band for this listing on Mercor; short tasks are typically paid against actual time logged and nothing here is guaranteed.
Downstream. The stated benefit is consideration for future clinical evaluation projects, which in practice is the main reason experienced clinicians take a 30-minute task.

What they look for

Graduate degree or professional qualification in a clinical mental health discipline (clinical/counselling psychology, psychiatry, clinical social work, psychotherapy, MFT, mental health counselling)
Direct practice experience in a clinical or behavioural health setting
Working familiarity with the real texture of clinician–client dialogue, including resistance, disclosure pacing, and rupture
Ability to make independent qualitative judgments without a rubric-by-committee
Attention to fine conversational detail — turn-taking, hedging, affect mismatch
Comfort working self-directed through a web evaluation interface

Included with this listing

Interview prep — 10 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether your degree and your hours are in a discipline the listing names, and whether the clinical exposure is direct rather than supervisory or academic.
A strong answer: Name the qualification, the awarding body, the setting type (community mental health, inpatient, private practice, IAPT-equivalent, school-based), the population, and roughly how many years or client hours of direct contact.
Where people lose it: Leading with research or teaching credentials and only mentioning clinical contact in passing — this task is built on the assumption you have sat across from clients, and a CV that reads academic invites follow-up you may not want.