What the work involves
You will spend most of your time reading model output and deciding whether it is psychologically sound — not just grammatically correct Gujarati. Typical tasks include assessing AI-generated responses to clinical vignettes for diagnostic accuracy and ethical risk, authoring original case scenarios in Gujarati that reflect how distress, family dynamics, and help-seeking actually present in Gujarati-speaking communities, and checking that an English assessment item survives translation with its construct intact. Documentation matters: your rationale is the training signal, so short, specific written justifications are part of every deliverable rather than an afterthought.
What the platform screens for
micro1 uses an AI-led interview before any human review. It probes two things hard. First, whether your Gujarati is genuinely working-professional rather than conversational — you may be asked to render terms like dissociation, rumination, or intrusive thought and explain why a literal translation would mislead a client. Second, whether you can criticise a plausible-sounding answer. Screeners often present output that reads fluently and is subtly wrong (a misapplied criterion, a safety-relevant omission, culturally imported assumptions about individual autonomy) and watch whether you name the specific defect or simply grade it favourably.
Logistics
- Fully remote contractor engagement, invoiced hourly; no clinical licence to practise is being used and you are not delivering care.
- Largely asynchronous with occasional scheduled calibration sessions; most contributors work 10–20 hours a week, though volume fluctuates with the customer's project phase.
- Expect an onboarding batch that is reviewed closely before steady-state volume opens up.
- Pay in the $100–200/hr band reflects rates observed on this category of micro1 engagement; the actual rate is set per project and is not guaranteed.