Sendou

Problem

For most people in the Central African Republic, seeing a doctor costs a whole day. You travel, you queue for hours, and only then do you find out whether the trip was worth it. Some make that journey for a fever that would have passed on its own. Others put it off until it is too late.

Doctors are not much better off. Many still keep their appointments in a paper book, with no secretary and no way of knowing who will actually turn up.

Sendou means "health" in Sango, the national language.

Solution

Sendou puts a first opinion and a doctor's diary in the patient's pocket. A medical assistant asks about the symptoms and tells the patient how urgent they look. From there, they book a doctor nearby and see them in person or by video, in French or Sango. Prescriptions, documents and reminders follow in the same app.

The assistant never makes a diagnosis. If the answers point to an emergency, it stops asking questions and sends the patient straight to emergency care.

The opportunity

The tools are already in people's hands: an Android phone and a mobile money account. What is missing is the link between a worried patient and an available doctor.

Sendou is built as a two-sided service, and that is what makes it a business rather than a public programme. Patients save the trip and the wait. Doctors get an organised diary, fewer missed appointments, and a summary of the case before the patient walks in. Consultations are paid in the app with Orange Money or Airtel Money, so a doctor's income is tracked from the first booking.

Leading it

I founded Sendou and run it as CEO, with a team of engineers, doctors and people who handle partnerships. Much of my work is keeping those three groups pulling in the same direction, because any one of them can sink the product on its own.

So we set a few rules early. Doctors own the medical side: nothing the assistant says changes without a doctor's sign-off. Engineers own the experience on a cheap phone over a weak connection, and the main screens have to load in a few seconds on 3G. And no personal data goes to an AI provider unless it has been anonymised first.

Hard calls

• Payments. Card payments are not an option for most of our patients, so we chose a local mobile money provider from day one, even though it meant more integration work.
• Sign-in. SMS codes are the obvious choice in a market of phone numbers, but they cost too much at launch. For now patients sign in with an e-mail link or with Google, and SMS will come back once an affordable provider is available in the country.
• Scope. Building for a 2 GB Android phone rules out features we would happily ship elsewhere. We chose reach over richness.

Next

The web app for doctors is finished. The patient app, the backend and the assistant are close behind, and mobile money payments are the last big piece before the first patients use it.