Regulatory medical writers author and grade evaluation tasks built from Clinical Study Reports, DSURs, and PSURs/PBRERs to train AI systems on clinical documentation. The work is judging whether AI-generated narrative meets ICH E3 structure and is genuinely supported by the underlying TFLs and protocol — and writing the rationale that explains why.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work involves

You write and review evaluation items drawn from the documents you already know: CSRs, DSURs, PSURs/PBRERs, and their supporting protocols and TFL packages. A typical task gives you a model-generated section — an efficacy summary, a safety narrative, a discussion, a conclusion — and asks you to decide whether it holds up. That means checking ICH E3 organisation and section sequencing, checking cross-references and appendix handling, and then doing the harder part: tracing each narrative claim back to the table, figure, listing, or protocol section that is supposed to support it. Where the model has overstated a result, silently dropped a subgroup, or attributed a finding to the wrong analysis population, you flag it and say exactly where the text and the source part company.

The written rationale is the deliverable, not a byproduct. Reviewers are graded on whether an engineer with no clinical background can read the rationale and understand what the model got wrong and what correct output would have looked like. A verdict without a traceable justification is treated as incomplete.

What the screen looks for

micro1's screening is AI-led and conversational, with follow-ups that go a layer deeper than the first answer. It is looking for people who have personally driven a full CSR to final — not contributed a section — and who can talk fluently about ICH E3 appendix structure, E2C/E2F periodic reporting conventions, and eCTD placement without hedging. Expect to be pressed on specifics: which TFL you would open to verify a given sentence, how you distinguish an interpretive error from a house-style preference, how you handle a claim that is technically supported but materially misleading. Oncology and haematology experience is weighted heavily.

Logistics

Fully remote, contractor engagement, invoiced hourly
Largely asynchronous task work with periodic calibration sessions with project leads
Volume varies by project phase; treat it as supplementary rather than full-time unless a lead says otherwise
Observed band is $30–60/hr, set by depth of authoring history and therapeutic-area fit — stated as observed, not guaranteed

Medical Writer (Pharmaceutical)

What the work involves

What the screen looks for

Logistics

Interview prep — 10 questions