Medicine
Annotation
Physicians — documentation feedback
Licensed physicians provide feedback to improve real-world healthcare documentation systems.
Fixed rate / hour
Apply
Medicine
Clinical evaluation of diagnostic AI reasoning
Help a medical AI get diagnosis right. Your clinical judgment trains systems that 200+ clinics will deploy next quarter. This is the kind of work that changes outcomes.
Zero fees
Set your hours
100% remote
Resume credential
This is the first time my medical expertise directly improves the AI tools I use in my own practice.
Who this is for
- Active medical license & general/primary care practice
- Comfortable with async, written evaluation work in English
- 10–20 hours/week for 4–8 weeks
Not for full-time seekers or those without clinical licenses
Not sure? Apply anyway — we review edge cases.
About this project
A US-based medical AI company is building a diagnostic reasoning system for primary care. They need licensed physicians to evaluate AI diagnostic outputs across 800–1,200 clinical scenarios — identifying reasoning errors, missed differentials, and miscalibrated confidence.
Your evaluations directly inform the next training cycle. The system is targeted for deployment in 200+ clinics next quarter.
What you'll do
- Review AI-generated diagnostic reasoning for clinical scenarios
- Flag errors, missed differentials, and confidence miscalibration
- Provide concise written feedback (1–3 sentences per case)
- Participate in weekly 30-min calibration calls with the research team
Qualifications
- Active medical license (MD, DO, or international equivalent)
- Minimum 5 years of clinical practice
- Conversational English (scenarios are in English)
Preferred
- Primary care specialty
- Familiarity with clinical decision-support tools or EHR systems
- Prior research or publishing in clinical AI
Adjacent expertise counts. Apply anyway.
Compensation & schedule
$85–110 per hour based on credentials and experience. Paid via PayPal, Payoneer, or bank transfer. No platform fees.
Flexible schedule: target 10–20 hours per week, but you set your own pace. Project runs 4–8 weeks.
FAQ
When and how will I be paid?
You're paid for the hours you log, sent via PayPal, Payoneer, or bank transfer. There are no platform fees — the rate you see is the rate you keep.
Do I have to work set hours?
No. Aim for 10–20 hours a week, but you choose exactly when you work. The project runs 4–8 weeks.
What happens after I apply?
We review your credentials and match you to the right tier, usually within a couple of days. If it's a fit, you get platform access and your first scenarios.
Is my work kept confidential?
Yes. Your evaluations are handled under strict confidentiality and used only to improve the model.
What if I'm not a perfect fit?
Apply anyway — we review edge cases, and adjacent clinical expertise often qualifies.
Similar projects
Generalist
Data Collection
Share online purchase emails — Andromeda
Share emails from online purchases to help build datasets used in AI development. Worldwide.
Per approved asset
Apply
Languages
Annotation
Audio annotation & speaker identification — Kawaii
Train speech recognition models by labeling conversational interactions in your native language.
Fixed rate / hour
Apply
OneForma Equal Opportunity
OneForma is an equal-opportunity employer. We will not discriminate based on race, gender, religion, or cultural background.