I Built a Doctor in a Box. It Runs Offline. It Could Save Lives in Africa.
In 2003 I received an NIH grant to build a clinical decision engine that could produce a medical differential diagnosis‚ and replicate the reasoning of any individual physician. It worked.
Twenty-three years later I picked the idea back up. This time with AI as my partner. The result is Occam Rx.
The Problem
A Peace Corps volunteer is stationed in a village in the Democratic Republic of Congo. No doctor within 80 kilometers. No internet. No electricity grid.
A child arrives with high fever, shaking chills, and confusion. Malaria? Meningitis? Typhoid? The symptoms overlap. Guessing wrong is fatal.
She has her training. She has her instincts. She does not have a tool built for this exact moment.
Until now.
What Occam Rx Does
Occam Rx is a fully offline clinical decision support app built on four tables and one scoring engine.
A volunteer walks a patient through 30 intake questions — fever pattern, respiratory symptoms, neurological signs, skin changes, exposure history. Each answer fires weighted scores against 20 diseases endemic to central and eastern Africa. The five highest-scoring diseases surface as the differential diagnosis in real time.
Not one answer. A ranked differential. Exactly how a good physician thinks.
It runs on a $150 Android tablet. Charged by a $35 solar power bank. Zero internet required.
What Makes It Different
Five languages - English, Kiswahili, French, Amharic, Lingala
Vital signs wired into the scoring engine‚ heart rate, temperature, SpO2, blood pressure, MUAC
HIV status and pregnancy - both feed the differential automatically
Field drug reference‚dosing by weight and age for all 20 diseases
Patient records and analytics** ‚every intake saved anonymously on device
Full patient chart, tap any record to see the complete clinical picture
The 20 diseases: Malaria. TB. HIV/AIDS. Typhoid. Cholera. Dysentery. Pneumonia. Meningitis. Dengue. Yellow Fever. Schistosomiasis. Sleeping Sickness. Leishmaniasis. Measles. Malnutrition. Anemia. Hepatitis B. Tetanus. Rabies. Intestinal Parasites.
Built With AI
I want to be transparent. I built Occam Rx in collaboration with AI writing the scoring engine, populating medical content, translating five languages, designing the UI.
This is what AI-assisted development looks like at its best. My clinical architecture expertise from 2003. Claude's medical knowledge and engineering capability. Neither of us could have built this alone.
The Mission
A Peace Corps worker in a remote village in central or eastern Africa, with no doctor, no internet, and a sick patient in front of them, picks up a small tablet and gets a clinically sound differential diagnosis with actionable field remedies in under 5 minutes.*
Matthew J. McCarthy ‚ Professor of Information Systems, W. P. Carey School of Business, Arizona State University. NIH grant recipient 2003.
@WPCareySchool @asu @WPCdean