Paul John

Paul John

Software Engineer

Listen to summary

MedFlow

AI • Healthcare • Health System

Overview

MedFlow started as an AI x Healthcare initiative focused on patient-flow inefficiencies in Rwanda. The goal was to design a system that helps patients get the right level of care faster while reducing avoidable pressure on urban hospitals.

Problem

Patients often flock to well-known hospitals regardless of the severity of their condition, causing massive bottlenecks. A patient might wait 6 hours just to see a doctor for a minor issue that could have been treated at home or at a less crowded local clinic.

Approach

To address this, we conceptualized MedFlow, a system that puts a basic pre-diagnosis step on the patient's device before they ever leave home. Instead of walking into a crowded hospital blind, you already know what you might be dealing with, where to go, and when to show up.

Implementation

The system is built around a triage-first, route-smart workflow:

  • Digital Triage: Users answer structured symptom questions on their mobile or web device. The system uses AI to score case urgency and produce a pre-diagnosis.
  • Smart Routing:
    • Non-Critical: You receive home-remedy guidance and an option to schedule an appointment if needed.
    • Critical/Serious: You are routed to the nearest health facility that can actually handle your case — not just the most popular one — with your pre-diagnosis attached to the request.
  • Appointment Scheduling: You book a specific time slot, so you know exactly when to go instead of spending half a day waiting.
  • Health System Integration: On the facility side, the system integrates with health systems to pull availability data, allow staff to respond to appointment requests, and receive pre-diagnosis information ahead of the patient's arrival.

Current State

MedFlow is currently at the conceptual and design-validation stage, with a defined workflow, PRD, and architecture direction. The core value proposition is clear: patients arrive faster, with better information, at the right facility — while underutilized facilities get the traffic they can handle.