AI Emergency Dispatch Intelligence

Closing the pre-hospital
gap-where most
lives are lost.

Sauti Health is an AI-powered emergency dispatch platform built for resource-limited settings. We turn existing community assets — trained first responders, motorcycle networks, and mobile infrastructure — into a coordinated, measurable emergency response system.

See How It Works Contact Us
5.8M
Preventable deaths from injury annually in low- and middle-income countries
Haagsma et al., Injury Prevention, 2016 · WHO Global Status Report, 2023
90%
Of global injury-related deaths occur in LMICs — where coordinated pre-hospital response is largely absent
GBD 2019 Disease and Injury Collaborators, The Lancet, 2020
10%
Emergency patients in sub-Saharan Africa transported by ambulance — 90% find their own way, or don't
Ningwa et al., BMC Health Services Research, 2020
$7.54
Cost per DALY averted in our pilot — 96.9% below the WHO cost-effectiveness threshold
Internal cost model · WHO GDP per capita threshold, 2022
01 — The Problem

A mathematically
unsolvable problem
by conventional means.

Building Western-model EMS in low-income settings requires $50–$280 per capita — figures that exceed annual health budgets entirely. Training enough paramedics takes decades. Emergency numbers exist in name only. This is not a funding gap to be closed incrementally. It requires a fundamentally different architecture.

5.8M
Preventable deaths annually in LMICs
90% of global injury-related deaths occur in low- and middle-income countries. The driver is not injury severity — it is the absence of any coordinated pre-hospital response.
Haagsma et al., Injury Prevention, 2016 · WHO Global Status Report, 2023
$50–280
Per-capita annual EMS cost — Western model
For a country of 45 million, this means $2.25B–$12.6B to replicate high-income EMS — often three or more years of the entire national health budget.
Ifeanyichi et al., BMC HSR, 2021 · Delaney et al., Surgery, 2024
50+ yrs
To close the paramedic workforce gap at current training rates
In our pilot market: 30 emergency physicians vs 5,580 needed; ~200 paramedics vs 45,000 needed. No incremental hiring strategy closes this in a policy-relevant timeframe.
MOH Uganda Workforce Assessment, 2024
60%+
Pre-hospital mortality rate across sub-Saharan Africa
Most deaths in emergency settings occur before hospital contact. Response time, not hospital quality, is the primary determinant of survival — and response time is structurally broken.
MOH Uganda Emergency Care Systems Assessment, 2024 · Ningwa et al., 2020
912
Allocated emergency number in Uganda — never operational
Inactive due to the absence of a dispatch system, per MOH 2022 report. Government has publicly acknowledged the gap and expressed openness to private sector solutions. This pattern repeats across the region.
Uganda Ministry of Health, 2022
30.8%
Ambulances with required equipment and trained staff
In our pilot market, 69% of ambulances are entirely unequipped for emergency care — functioning as patient taxis. Equipment gaps compound the workforce gap at every level.
Ningwa et al., BMC Health Services Research, 2020
02 — Approach

Coordinate what exists.
Don't wait for what doesn't.

We don't replace ambulance systems — we make them unnecessary for the first critical window. Sauti Health activates the infrastructure that already exists in every resource-limited setting: community health workers, motorcycle networks, and near-universal mobile penetration.

Cost Benchmark — Per Emergency Response

The cost gap between Western EMS and community-based dispatch is not a margin — it is an order of magnitude. Sauti Health operates at a fraction of the cost of the minimum viable ambulance-based system, using infrastructure that already exists.

Cost Comparison — Per Emergency Response
Western EMS — full deployment up to $280
Minimum ambulance model — LMICs $50+
Sauti Health — deployment phase ~$1.60
Reduction vs minimum ambulance model ~97%
Ifeanyichi et al., BMC HSR, 2021 · Internal cost model
Infrastructure Coordinated — Uganda Pilot
CHW
287,000+
Community Health Workers (VHTs)
WHO BEC-trained first responders · 62% active rate · 20 years embedded in communities · deployable within minutes
MX
800,000+
Motorcycle Taxi Riders (Boda Boda)
142,000 in Kampala alone · trackable via SafeBoda, Uber Boda, Bolt · 95%+ with mobile money accounts
4G
75%
Mobile Penetration — Uganda
98% 4G on pilot corridor · USSD works on 2G feature phones · zero data, zero smartphone required
MM
65%
Mobile Money Adoption
MTN MoMo + Airtel Money enable instant per-incident payment to responders. No cash. No delay.
03 — Technology

An EMS digital twin.
Built for feature phones,
2G, and mobile money.

Five integrated layers coordinate community responders in real time. The system operates offline for up to 72 hours, dispatches via USSD on any 2G phone, and learns continuously from every emergency handled. Each layer is purpose-built for the infrastructure realities of resource-limited settings.

01
Data Ingestion & Access
USSD on any phone, any network, 2G minimum — 100% population reachable without smartphones. Offline-capable app for responders with smartphones. IVR for low-literacy or hands-free callers. 72-hour offline sync queue with automatic restoration. Patient data encrypted at rest (AES-256) and compliant with local data protection legislation.
USSD · IVR · Offline
02
Dispatch Algorithm
Deployment phase: geographic zone-based matching with 500m corridor segmentation. Cascade logic: if no acceptance within 90 seconds, automatically escalates to the next zone. Scale phase: full AI dispatch using Deep Q-Network (DQN) trained over 50,000 episodes with multi-objective reward — response time, severity match, geographic equity, resource efficiency.
Zone dispatch · DQN · AI
03
Mobile Money Payment Engine
MTN MoMo + Airtel Money APIs integrated for instant, automatic responder payment on incident completion. Retry logic at 5, 30, and 60 minutes before manual flag. Every transaction logged with timestamp, recipient, amount, and error code. No responder waits more than 24 hours for payment. Designed to port to any mobile money infrastructure.
MTN MoMo · Airtel · API
04
Bayesian Calibration & Continuous Learning
Hamiltonian Monte Carlo (HMC) via PyMC implements Bayesian hierarchical modelling to account for sub-district heterogeneity. The digital twin continuously updates parameter posterior distributions as real field data streams in. Simulation trained on 500 replications × 100,800 minutes — equivalent to 137 simulated years of operations.
PyMC · HMC · Bayesian
05
Operations Dashboard & Analytics
Mobile-first coordinator dashboard: live active incidents, elapsed time, non-response flags. Running totals: incidents to date, average response time, payment success rate. Demand heatmaps by zone and time-of-day. DHIS2 integration for population health data overlay. Scenario planning tools for capacity expansion and health systems partner presentations.
Streamlit · DHIS2 · Plotly
04 — Health Impact

DALYs averted.
Three scenarios.
Full methodology shown.

Delayed pre-hospital response produces a measurable burden of disability and premature death. We calculate DALYs averted using the standard WHO/GBD methodology: DALYs = Years of Life Lost (YLL) + Years Lived with Disability (YLD). Figures are drawn from our Uganda pilot. All assumptions are documented transparently.

Input Parameters & Sources — Uganda Pilot
DALY Calculation Parameters
Life expectancy at birth (pilot market)
63.0 years
WHO Global Health Observatory, 2022
Mean age of emergency patient
28 years (RTI dominant)
Uganda Police Traffic Division, 2023
YLL per prevented death
63 − 28 = 35 years
Standard GBD methodology
Baseline pre-hospital mortality
60% RTI · 70% obstetric
MOH Uganda, 2024 · Ningwa et al., 2020
Simulated mortality reduction
15% (conservative) → 25%
Henry & Reingold, J Trauma, 2012 · Simulation
Disability weight (RTI non-fatal)
0.368 (moderate, 6 months)
GBD 2019 disability weights
Ratio non-fatal : fatal
~4:1 (4 survive per death)
Uganda RTI surveillance data, 2023
$7.54
Cost per DALY averted at initial scale (conservative scenario)
$243
WHO cost-effectiveness threshold (pilot market GDP per capita, 2022)
96.9%
Below WHO threshold — highly cost-effective even in the conservative scenario
~$51
Cost per DALY at scale (~3,000 emergencies/yr) — still well below threshold
DALY Calculation — Three Scenarios
Formula: DALYs averted = (Deaths prevented × YLL) + (Disabilities prevented × disability weight × duration)
Conservative — 15% mortality reduction
RTI deaths prevented: 60 × 60% × 15% = 5.4 deaths YLL: 5.4 × 35 yrs = 189 life-years Non-fatal YLD: 60 × 40% × 4:1 ratio × 15% × 0.184 = 2.6 YLD Obstetric deaths prevented: 20 × 70% × 15% = 2.1 → 73.5 YLL Total DALYs averted (conservative): ~265 DALYs (100 incidents) At scale (~3,000 emergencies): ~7,950 DALYs/yr
100 incidents: ~265 DALYs  ·  Scale: ~7,950 DALYs/yr
Moderate — 20% mortality reduction
RTI deaths prevented: 60 × 60% × 20% = 7.2 deaths → 252 YLL Non-fatal YLD: 60 × 40% × 20% × 0.184 = 0.9 YLD Obstetric: 20 × 70% × 20% = 2.8 deaths → 98 YLL Total DALYs averted (moderate): ~351 DALYs At scale: ~3,000 emergencies/yr → ~10,530 DALYs/yr
100 incidents: ~351 DALYs  ·  Scale: ~10,530 DALYs/yr
Optimistic — 25% mortality reduction
RTI deaths prevented: 60 × 60% × 25% = 9 deaths → 315 YLL Non-fatal YLD: 60 × 40% × 25% × 0.184 = 1.1 YLD Obstetric: 20 × 70% × 25% = 3.5 deaths → 122.5 YLL Total DALYs averted (optimistic): ~439 DALYs At scale: ~3,000 emergencies/yr → ~13,170 DALYs/yr
100 incidents: ~439 DALYs  ·  Scale: ~13,170 DALYs/yr
Methodological notes: Calculations use standard GBD DALY methodology without age-weighting or discounting, consistent with current WHO practice. Mortality reduction percentages are derived from simulation outputs calibrated to trauma literature (Henry & Reingold, J Trauma, 2012). All three scenarios are presented because honest uncertainty is more useful than a single point estimate. Full methodology available in the research protocol.
05 — Team

Built from evidence,
grounded in field reality.

Our founding team combines field implementation knowledge, emergency medicine expertise, and technology development — each with specific, verified experience relevant to the problem.

HT
Hillary Turinawe
Founder & CEO
Medical Student · University of Bonn, Germany
Leads strategy, fundraising, and institutional partnerships across Sub-Saharan Africa and Europe.
RK
Ronald Kyeyune
Head of Mobilisation & Partnerships
Community Health Mobiliser · Kampala, Uganda
Drives field mobilisation, responder recruitment, and B2B institutional partnerships on the ground.
OJ
Onama John
CTO & Head of Product Development
Full-Stack Engineer · USSD · IVR · Mobile
Builds the dispatch platform — from IVR ingestion and USSD routing to the AI engine and analytics dashboard.
BM
Dr. Benard Mwesigye
Founding Medical Director
Emergency Medicine Physician · Mbarara University of Science and Technology
Leads clinical protocol design, ethics clearance, and peer-reviewed research publication.
JB
Dr. Jacob Busingye
Founding Medical Director
Emergency Medicine Physician · Uganda
Pre-hospital care clinical governance and field protocol validation across the pilot corridor.
AV
Academic & Clinical Advisors
Research · Emergency Medicine · Global Health
Uganda · Germany · International
Advisory network spanning emergency medicine faculty, public health researchers, and global health technologists. Full panel available on request.
06 — Contact

Let's talk about
what's possible.

We welcome conversations with researchers, health systems partners, technology collaborators, and organisations interested in scaling emergency care in resource-limited settings.

@
Email
hillaryturin@gmail.com
☏
Phone
+49 157 3785 8903
↗
Website
sautihealth.com
Operational Base
Kampala, Uganda
Northern Bypass Corridor, Kampala Metropolitan Area.
Active field operations, community partnerships, and pilot coordination.
Bonn, Germany
European research and EMS collaboration base.
Academic partnerships and technology development.