What the work actually involves

You will spend most of your time on three kinds of tasks. First, writing original psychological material in Marathi and English — case vignettes, assessment items, intake dialogues, scenarios that a model can learn from — grounded in how distress, family structure, stigma and help-seeking actually present in Maharashtra rather than in a translated American textbook. Second, evaluating AI-generated responses to psychological prompts: is the clinical reasoning sound, is the crisis handling appropriate, does the Marathi read like something a Marathi speaker would say, does it overstep into diagnosis or advice a model should not give. Third, documenting your reasoning so another reviewer can reconstruct why you scored something the way you did.

The translation component is the part most candidates underestimate. Much of clinical vocabulary has no settled Marathi equivalent, and the colloquial terms people actually use for symptoms carry connotations that formal registers strip out. Work here often means deciding which rendering preserves clinical meaning versus which preserves how a patient would speak — and writing down why.

What the screen looks for

micro1's screening is AI-led and conversational. It probes whether the PhD is real and where it was earned, whether clinical hours exist alongside the academic record, and — most consistently — whether Marathi fluency extends to technical and clinical registers rather than household conversation. Expect to be asked to explain a construct like dissociation or anhedonia in Marathi and then justify your word choice. Expect follow-ups that push on ethics: what you would flag, what you would refuse to write, how you would handle a model output that gives plausible-sounding but harmful guidance. No AI or annotation background is required, and saying you have none costs nothing.

Logistics

  • Remote, contractor engagement, invoiced hourly; observed band for this posting is $100–200/hr, set by credentials and project, not guaranteed.
  • Largely asynchronous with occasional calls to interdisciplinary teams; hours are flexible but projects usually want a stated weekly commitment you can hold for several weeks.
  • Work happens in a browser-based annotation environment — familiarity with remote collaboration and documentation tools is expected, though the tooling itself is taught.