AI-Powered Decision Support to Help Public Ambulance Services in Kenya Deliver Faster, Safer Emergency Medical Care
Share project
Location
Project size
Budget
Requested services
Overview
An integrated AI system that supports ambulance dispatchers, EMTs, and supervisors with real-time triage, clinical guidance, and operational insights to improve emergency response outcomes.
The problem
Problem statement
Ambulensi supports public ambulance dispatch and prehospital emergency care across Kenya, where services are predominantly delivered by non-specialist dispatchers and EMTs operating in highly manual environments. While Ambulensi has successfully digitised ambulance coordination, most frontline users lack formal emergency medical dispatch (EMD) or advanced prehospital training. Dispatchers must assess emergency calls, determine priority, allocate ambulances, and guide callers through lifesaving actions with limited structured triage tools and no real-time decision support. Similarly, EMTs and paramedics deliver care under extreme time pressure while manually referencing static PDF-based protocols and SOPs, which is slow, cognitively demanding, and prone to error. As emergency call volumes increase and services scale across counties, reliance on manual judgement, static documents, and variable experience leads to delayed or inappropriate dispatch prioritisation, inconsistent quality of care, inefficient use of scarce ambulance resources, and preventable morbidity and mortality. The core challenge is not the absence of technology, but the absence of embedded intelligence within the system to guide users with varying levels of experience. Without AI-enabled decision support, Ambulensi risks replicating manual processes digitally rather than transforming emergency response outcomes.
The project will integrate directly into Ambulensi, EMKF’s existing, production-grade ambulance services platform. AI components will be embedded within the live web-based dispatch system, mobile applications for ambulance crews, and real-time analytics dashboards—rather than developed as a standalone tool.
What's already in place
hosting_domain, content, datasets
Team
The project will be led by EMKF’s in-house technical team, consisting of three software developers with experience building, maintaining, and scaling the Ambulensi platform. The team has strong expertise in full-stack development, API integrations, cloud infrastructure (AWS and Google Cloud), GPS-based systems, and health data analytics. They will be responsible for implementing the AI features, integrating them into existing dispatch and clinical workflows, testing functionality in live environments, and supporting iterative improvements during pilot and scale-up phases.
Project scope
AI-Assisted Dispatch Triage and Prioritisation Design and implement AI models to analyse dispatcher-entered caller information and generate real-time recommendations for priority, response level, and pre-arrival actions. Expertise needed: emergency triage logic, clinical AI design, model validation in low-resource settings.
Pre-Arrival Instruction Engine Convert existing PDF-based protocols into structured, machine-readable workflows and build an AI prompting layer that delivers safe, concise, context-appropriate instructions usable by non-clinical dispatchers. Expertise needed: protocol digitisation, human-centred AI design, clinical safety guardrails.
Field-Level Clinical Decision Support for EMTs Integrate SOPs into AI-guided care pathways with voice-assisted prompts and offline functionality. Expertise needed: applied GenAI in clinical workflows, speech interfaces, safety-critical system design.
Intelligent System Optimisation Develop AI models for demand forecasting, ambulance load prediction, and system optimisation using historical dispatch and geospatial data. Expertise needed: predictive analytics, spatiotemporal modelling, health systems optimisation.
Data Engineering and Model Governance
Users and functionalities
Primary Users
Emergency Medical Dispatchers (EMDs) Often non-specialist staff responsible for receiving emergency calls, entering caller information, prioritising cases, and coordinating ambulance response.
Ambulance Drivers and EMTs/Paramedics Frontline providers delivering prehospital care under time pressure, frequently with limited access to real-time clinical guidance.
County EMS and Health Managers Oversight users responsible for monitoring performance, resource utilisation, and system accountability.
Core AI-Enabled Functionalities
AI-Assisted Dispatch Triage and Prioritisation Dispatchers type structured information based on what callers report (symptoms, location, mechanism of injury, patient status). AI analyses these inputs in real time to recommend dispatch priority, response level, and next steps, supporting consistent decision-making regardless of dispatcher experience.
Pre-Arrival Instruction Support Based on dispatcher-entered information, the system prompts protocol-aligned, lifesaving pre-arrival instructions (e.g. bleeding control, positioning, basic CPR) for dispatchers to relay to callers while help is en route.
Intelligent Ambulance Allocation AI supports optimal ambulance assignment by considering urgency, proximity, availability, and system load, improving response times and fleet utilisation.
Voice-Guided Clinical Decision Support for EMTs Ambulensi’s existing SOPs and clinical protocols are transformed into AI-guided, step-by-step prompts delivered through the mobile app, reducing reliance on static PDFs and minimising errors under pressure.
Operational Analytics and Decision Support AI-enhanced dashboards surface performance trends, bottlenecks, and system risks to support data-driven planning, governance, and accountability at the county level.
Success vision
AI integration within Ambulensi will significantly improve both the quality and reach of emergency medical services for communities served by public ambulance systems in Kenya.
In terms of quality, patients and callers will receive faster, safer, and more appropriate responses. AI-supported triage will improve prioritisation of high-risk emergencies such as obstetric complications, trauma, cardiac events, and critically ill children. Pre-arrival instruction support will enable callers to begin lifesaving actions immediately, improving survival and reducing complications. In the field, AI-guided clinical support will standardise care delivery, reduce errors under pressure, and improve continuity of care during transport, especially in rural and underserved settings.
In terms of reach, Ambulensi currently serves over 6.2 million people. By improving efficiency, prioritisation, and resource utilisation, AI will enable the same ambulance fleets and workforce to safely handle higher call volumes, enabling expansion into additional counties and reaching tens of millions of Kenyans nationwide. Overall, the project will result in more lives saved, fewer preventable deaths, and more equitable access to timely emergency care.
Emergency Medicine Kenya Foundation
We are an NGO supporting governments and emergency healthcare providers across Kenya to save lives by strengthening the emergency healthcare system through capacity building, knowledge development, advocacy and research.
Website