Clinical Mental Health Expert: AI Conversation Evaluation

What you'll do

This is a short task, not an ongoing role. The ad says it takes up to 30 minutes to complete.

You are shown pairs of simulated conversations set in clinical mental health scenarios, and for each pair you decide which one more closely resembles a real clinical interaction. The judgment is about realism and authenticity: does the exchange sound like how a session with a patient or client actually goes, or does it read like a script? You do it on your own, through an online interface, using your clinical experience rather than a detailed rubric (the ad mentions none).

The purpose, per the ad, is an AI research initiative evaluating the quality of simulated clinical conversations. In practice that means your preferences help researchers tell which generated dialogues are convincing, which is the kind of signal used to improve how AI systems model therapeutic conversations.

Who qualifies

• A relevant degree or professional qualification in a mental health field
• Practical experience in a clinical mental health or behavioural health setting
• Familiarity with real conversations between mental health professionals and patients or clients
• Sound professional judgment about whether a clinical interaction feels real
• Attention to detail and the ability to make independent qualitative calls

The ad lists suitable backgrounds as clinical psychology, counselling psychology, mental health counselling, clinical social work, psychiatry, psychotherapy, behavioural health, marriage and family therapy, and other closely related clinical disciplines. That is broad: a practising counsellor or clinical social worker qualifies on the same footing as a psychiatrist.

No licence, minimum years of experience or country is specified. The ad does not name a working language; it is written in English, so expect the conversations to be in English.

What it pays

$100 per hour, Mercor's published figure. Because the task is capped at around 30 minutes, the headline rate works out to a small one-off payment. The ad does not say whether pay is calculated on actual time spent or a fixed task length.

The contract terms in the ad are Mercor's standard ones: independent contractor, weekly payment via Stripe or Wise based on services rendered, and no support for H-1B or STEM OPT candidates.

Worth knowing

Good:

• A genuinely small commitment: one sitting of up to half an hour
• A high hourly rate for work that draws directly on clinical experience
• Broad eligibility across counselling, social work, psychology, psychiatry and family therapy
• No patient contact and no documentation burden
• The ad says participants may be considered for future AI evaluation and research work needing clinical mental health expertise

Less good:

• At up to 30 minutes, the total earned is small; weigh it against the time Mercor's application and onboarding take
• Future work is a possibility the ad mentions, not a commitment
• H-1B and STEM OPT holders are excluded
• The ad does not say how many pairs you review, what the lab is, or how the conversations were generated

About this listing

Posted by Mercor as a remote hourly contract, read on 30 September 2026. The rate, the 30-minute duration, the backgrounds and the payment terms above come from the ad itself. No closing date or country requirement is published. See Mercor.