What the work involves

You write and critique psychiatric case material rather than treat patients. A typical task set might be: construct a case vignette of a patient presenting with somatic complaints masking depression in a Lithuanian primary care setting, specify the assessment path a local psychiatrist would actually follow, then judge whether a model's proposed differential, risk assessment, and next steps hold up. Other tasks run the other direction — you review model-generated Lithuanian mental health text and mark where it is clinically wrong, where it is clinically fine but culturally implausible, and where the phrasing would land badly with a Lithuanian patient or family.

Because the project is bilingual, much of the value sits in the gap between languages. Diagnostic vocabulary, ICD-10 coding habits, stigma dynamics, family involvement norms, and the way risk is documented do not translate cleanly from English-language literature into Lithuanian practice. You are expected to make those differences explicit in writing, not smooth them over.

What the screen looks for

micro1's screening is AI-led and follows up on whatever you say. It checks the credential chain first — medical degree, completed psychiatry residency, valid licence — then pushes into clinical specifics: how you actually structure a suicide risk assessment, what you do with a first-episode psychosis presentation, how you document capacity. Expect it to test Lithuanian fluency at professional register, not conversational level, and to test English well enough to write analysis a non-clinical AI team can use. It also probes ethical judgment on sensitive content: you will be reviewing self-harm, abuse, and crisis material, and the screen wants evidence you can handle it as review work under confidentiality norms rather than as clinical intervention.

Logistics

  • Remote contractor engagement, invoiced hourly; observed band for this role is $100–200/hr, set by project and seniority rather than guaranteed.
  • Largely asynchronous task work with periodic calls with AI project teams; most contributors commit a defined weekly block (often 10–20 hours) around clinical duties.
  • No prior AI or annotation experience is required or expected.
  • Written output is required in both Lithuanian and English, so plan for the language work to take real time.