What the work involves

You review psychological material — case vignettes, assessment write-ups, clinical reasoning chains, model responses to user distress — and judge whether it is accurate, ethically sound, and consistent with professional practice. A second strand is generative: writing Georgian-language scenarios, prompts, and reference answers that reflect how psychological concepts actually land in a Georgian cultural and family context, rather than reading as translated American clinical text. Expect to justify each judgment in writing, with enough specificity that another psychologist could reproduce your reasoning.

Bilingual fidelity is the core technical demand. Terms like distress, boundaries, attachment, or trauma carry different weight in Georgian clinical usage than in English, and a chunk of the work is deciding when a literal rendering misleads and what to substitute. You will also flag content that is clinically plausible but ethically wrong — a model that gives a competent-sounding formulation while missing a duty-to-warn trigger, or that pathologises ordinary family structure.

What the platform screens for

micro1 runs an AI-led interview before human review. It checks that the PhD is real and where it was earned, probes clinical and academic experience with follow-ups, and tests Georgian depth on technical vocabulary rather than conversational ability. Expect scenario prompts: you are shown a flawed AI answer and asked what is wrong with it and how you would rank it against an alternative. Reasoning that stays concrete and cites the standard or the clinical rationale scores well; confident generalities do not. No AI or annotation background is required.

Logistics

  • Fully remote, contractor engagement, asynchronous with occasional interdisciplinary calls
  • Hours are flexible and project-driven; many contributors run 10–20 hrs/week, with volume varying by project phase
  • Pay observed in the $100–200/hr range, not guaranteed; rates depend on project, seniority, and task type
  • Reliable internet and comfort with browser-based annotation and documentation tools expected