Write hard, original medical Q&A pairs that stump frontier AI, source and cite the answers from primary literature and guidelines, and make questions harder when models get them right. Open to medical students, residents and physicians. Paid per accepted task with a weekly minimum. Five openings, $40–90/hour.
Write and review evaluation tasks from CSRs, DSURs and PSURs/PBRERs, and judge AI-drafted clinical documents on ICH E3 structure, scientific accuracy and whether every claim traces to the TFLs. Five years of regulatory writing with full CSR authorship. Thirty openings, contractor, $30–60/hour.
Write multi-step benchmark questions on clinical and epidemiological charts (Kaplan-Meier curves, forest plots, ROC curves, hazard-ratio plots) with unambiguous, fully worked answers for AI evaluation. MD, MPH, PhD or related degree plus two years. 50 openings, contractor, $25–50/hour.
Write and review evaluation tasks from CSRs, safety reports and regulatory correspondence in oncology and haematology, judging efficacy and safety conclusions, dose escalation, AE grading and benefit-risk reasoning. Board-certified oncologists or haematologists with five years post-training and trial experience. Thirty openings, $160–200/hour.
Write and review evaluation tasks from MedEd modules, advisory board summaries, slide decks and HCP materials, and judge AI-generated MedComms for accuracy, fair balance, audience framing and source fidelity. Five years in medical communications, MedEd or medical affairs. Fifty openings, contractor, $50–80/hour.
Review inpatient cases for clinical quality, documentation accuracy and appropriate care, make the call on unclear or unusual cases, and help shape the review standards used to train AI. MD or DO with ten years of attending hospitalist experience. Thirty openings, contractor, $70–100/hour.
Write and review evaluation tasks that make an AI derive, reproduce or validate clinical trial statistics from TFLs and CDISC datasets, and catch mismatches between outputs, the SAP and the CSR narrative. Remote contractor, $60–65/hour, 30 openings; 5+ years in clinical trials and SAS or R preferred.
Assess adolescent eating disorder cases in writing, applying DSM-5, EDE-Q and SCOFF, judging medical stability against MEED, and saying whether FBT or CBT-E fits the clinical stage. Wants a licensed clinician with a recent adolescent caseload, and accepts paediatric medicine alongside psychiatry and psychology. Remote contractor work.
Review and annotate data sets on children's learning and behaviour, judge model output for developmental and cultural accuracy, and write the guidelines that shape how an AI handles the linguistic, cognitive and psychosocial sides of child development. A PhD in the field is required; research or academic background counts, with or without clinical experience.
Review case studies and simulated patient interactions for military and veteran populations, then write the expert feedback, prompts and scenarios that teach a model what good clinical practice looks like, including the cultural nuance this population expects. Asks for a completed psychiatry residency or a psychology PhD, plus direct patient care experience.
Assess adolescent substance use case material, stage severity against DSM-5 SUD and ASAM criteria, and recommend a level of care, separating developmentally normal experimentation from clinically significant patterns, and flagging withdrawal, overdose and safeguarding risks. Wants a licensed clinician with a recent adolescent caseload, not AI experience.
Read adolescent case material and write structured suicide and self-harm risk assessments, applying C-SSRS and formal risk formulation, separating ideation from intent, and recommending escalation pathways and safety planning. The output trains models meant to guide non-specialists. Wants a licensed clinician with a recent adolescent caseload, not AI experience.
Design and review physician and patient questionnaires (screeners, question wording, response scales, branching logic, respondent burden) and say whether an instrument will actually produce usable data. A flat $140/hour, and therapeutic-area specialism is welcome but not required. Fully remote, on your own schedule.
Read physician and patient research surveys and say whether the clinical terminology, treatment pathways and response options match how the disease is actually treated. A flat $200/hour, three years in a therapeutic area, and an MD is explicitly not required. Fully remote, on your own schedule.
New roles land most days. Get them on Telegram or Discord as they are added, or read how the platforms pay before you apply.