For medical professionals
Get paid to review AI's medical work.
Healthcare Reviewer matches verified physicians, nurses, pharmacists, and clinical researchers with flexible remote work reviewing medical AI: grading answers, catching unsafe guidance, and strengthening the tools patients will rely on.
$40–$200+
per hour, by credential tier
18
clinical specialties
100%
remote, no minimum hours
Weekly
payouts for approved work
Why it matters
This is patient-safety work.
Medical AI fails in ways only clinicians catch: a confident answer citing a trial that doesn't exist, a drug interaction invented from pattern-matching, a triage bot reassuring the one patient who needed an ambulance. Catching those failures takes people who have carried real responsibility for patients. That is why every reviewer on this platform is credential-verified before their first case.
How it works
Three steps from application to first payout
No recruiters, no interview loops. Verification and one qualification exercise stand between you and paid review work.
01
Apply & verify
Tell us your role, specialty, and license details. We verify your credentials against state boards, then you complete one short specialty-relevant qualification exercise.
~30-min exercise · verification in 2–3 business days
02
Get matched
We match you to projects by specialty and credential tier: an EM physician gets triage conversations, a PharmD gets interaction QA. You choose which to accept.
Specialty-based matching
03
Work & get paid
Review cases asynchronously in the platform, at whatever volume fits your week. Approved work pays out weekly by direct deposit.
Weekly payouts · no minimum hours
Why Healthcare Reviewer
Work that takes your training seriously
01
Rigor
Clinical-grade review, not busywork
You grade AI outputs against real standards of care: differentials, dosing, escalation decisions. The work assumes you think in rubrics and evidence, because you do.
02
Impact
Your catches protect patients
Every hallucinated interaction you flag and every unsafe reassurance you overturn makes the next model generation safer for the people who will eventually rely on it.
03
Flexibility
Fully remote, fully asynchronous
Work between shifts, post-call, or on research days. No meetings, no minimum hours, no schedule to protect. Most projects run 4–15 hours a week, at whatever pace fits your life in medicine.
04
Rate
Pay that respects your training
Rates reflect what clinical time is worth, up to $200+/hr for board-certified specialists. Every listing shows its rate before you apply.
The work itself
One case through the review loop
Every project is a variation on the same clinically honest loop: read, grade, correct.
01
Read the AI's work
A triage conversation, a drafted patient answer, a labeled clinical note. You see exactly what the model produced, on de-identified or synthetic cases.
02
Grade it and flag what's wrong
Score the output against a structured rubric and tag the failures: the missed red flag, the fabricated citation, the interaction that doesn't exist.
03
Write the correction
Author the answer the model should have given. Your corrected version becomes the training signal. This is where your clinical judgment does its work.
Clinical domains
Where models need clinicians most
Review work spans 18 specialties. These are the domains where expert judgment changes model behavior fastest.
Cardiology & Critical Care
Models still misread tracings and mishandle acuity. Reviewers grade ECG interpretations and escalation logic.
Pharmacology & Drug Safety
Hallucinated interactions are among the most dangerous AI failures. PharmDs verify every claim against primary references.
Diagnostic Reasoning & Imaging
Differential quality separates useful models from risky ones. Physicians grade localization logic and can't-miss coverage.
Emergency & Acute Care
Triage is where wrong answers cost the most. EM clinicians review acuity assignment and red-flag recognition.
Oncology & Pathology
Staging, biomarkers, and guideline pathways change faster than models retrain. Specialists catch what's outdated before it reaches a patient.
Mental & Behavioral Health
Patient-facing tools must never miss a crisis signal. Psychiatric clinicians red-team the conversations that matter most.
Who we verify
Built for every stage of a clinical career
If you think in differentials, dosing, or evidence tables, there's review work matched to your credential tier.
Medical students
$40–$75/hrMS1–MS4 at accredited MD/DO programs.
- Grade board-style Q&A for factual accuracy
- Verify guideline and trial citations
- Label de-identified clinical text
Residents & fellows
$65–$120/hrEvery specialty, PGY-1 through fellowship.
- Evaluate differentials and workup sequencing
- Review triage conversations for missed red flags
- Author reference answers for failed cases
Attendings & specialists
$100–$200+/hrBoard-certified physicians in active practice.
- Adjudicate specialty model evaluations
- Set gold standards junior reviewers grade against
- Red-team high-stakes clinical tools
RNs, NPs & PAs
$40–$110/hrRNs $40–$75/hr; NPs and PAs $60–$110/hr.
- Label nursing documentation and clinical notes
- Evaluate AI visit summaries against transcripts
- Ground product decisions in real workflows
Pharmacists
$60–$110/hrPharmDs and RPhs, inpatient or retail.
- Catch hallucinated drug interactions
- Verify renal and hepatic dose adjustments
- QA pediatric weight-based dosing
Clinical researchers
$60–$110/hrPhDs, MD-PhDs, and research staff.
- Trace AI citations to their actual sources
- Check statistic fidelity against published papers
- Build verified-claim evaluation datasets
Open now
Current opportunities
Clinical Documentation Workflow Survey (Nursing)
$45–$65/hr
Structured interviews and surveys about real documentation workflows to ground AI product decisions in how clinical work actually happens.
Medical Q&A Benchmark Grading (Internal Medicine)
$80–$120/hr
Grade AI answers to USMLE-style and clinical-reasoning questions: factual accuracy, reasoning quality, and appropriate uncertainty.
Radiology Report Annotation
$115–$160/hr
Annotate de-identified radiology reports for findings, impressions, and follow-up recommendations to train structured-reporting models.
FAQ
Common questions
Eligibility, verification, pay, and privacy, answered directly.
Your clinical judgment, applied at model scale
Verification takes minutes to start and 2–3 business days to complete. Browse open projects while you wait.