What the work involves
micro1 recruits credentialed medical writers to produce reference documents and to judge model output against professional standards. Day to day, tasks fall into a few recurring shapes:
- Authoring gold-standard drafts: a clinical study protocol synopsis, a CSR section, an informed consent form, a plain-language summary, or a manuscript abstract from supplied data.
- Reviewing model-generated clinical text for factual accuracy, citation fidelity, ICH E3/E6 and CONSORT alignment, and reading-level appropriateness for the stated audience.
- Ranking two or more model responses and writing a short rationale that a non-clinician reviewer can follow.
- Flagging failure modes that matter in this domain: fabricated references, misstated endpoints, off-label implications, unsupported causal language, and unsafe patient-facing phrasing.
The written rationale usually matters more than the score. Reviewers who explain why a sentence would fail a regulatory or journal review are the ones who get repeat work.
What the platform screens for
Expect an AI-led interview that probes specifics rather than credentials in the abstract. Interviewers ask what documents you have personally authored, at what phase, for which submission, and what changed after review. Vague claims tend to collapse under follow-up. Screening also tests whether you can distinguish a stylistic preference from a substantive error, and whether you can hold a consistent standard across several similar samples. A short writing or editing sample is common.
Logistics
Fully remote and asynchronous, with work delivered through the platform's task interface. Most contributors take 5–20 hours per week in self-selected blocks; some projects have batch deadlines measured in days rather than hours. Engagements are contract-based and volume fluctuates with client demand, so treat this as supplemental rather than a replacement for a primary role. Pay in the $50–80/hr range has been observed for this category; actual offers vary by project, specialty, and assessed calibration quality.