Computer vision, demand forecasting, and cold-chain telemetry — engineered for health supply chain teams operating without margin for error. Built for the field, integrated with the systems you already run.
2.41M
Doses tracked
1,240
Facilities
98.7%
Chain uptime
11
Countries live
VVM Scanner · live
Kano State warehouse W-03
Safe · Stage 1
34 vials
Watch · Stage 2
05 vials
Discard · Stage 3+
01 vial
SCAN HISTORY · LAST 4 H
Live VVM scanner readout from a partner facility. ↓ Scroll for more.
Integrates with
DHIS2
mSupply
OpenLMIS
FHIR R4
WHO PQS
UNICEF SD
Each module ships independently and integrates through the same data layer. Adopt one, adopt all — your existing LMIS keeps working underneath.
01
1,240 facilities online
Real-time inventory
Live stock levels across every facility, colour-coded as adequate, low, or critical. Filter by vaccine, country, or facility — spot shortfalls before they become stockouts.
Explore inventory →02
12-week horizon · 80% CI
AI-powered forecasting
A Prophet + LightGBM ensemble predicts demand surges, flags expiring stock, and models what-if scenarios per facility. Order the right quantity at the right time.
View forecasts →03
98.7% chain uptime
Cold chain monitoring
Continuous temperature telemetry with configurable thresholds, breach timelines, and excursion alerts routed to the right responder.
See cold chain →04
<100 ms inference
AR stock counter
Point a phone camera at a shelf — YOLOv8 counts vials, syringes, cold boxes, and ancillaries in real time and reconciles against the ledger.
Try AR scanner →05
4-stage classification
VVM stage classifier
A computer-vision model reads vaccine vial monitor indicators and flags heat-exposed vials (Stage 1 through 4) at the point of dispensing.
Explore vision AI →06
11 country programmes
Coverage intelligence
Interactive geospatial maps showing immunisation coverage by district and facility. Drill from country to clinic in a few clicks.
View coverage map →N ↑ · 1:4,200,000
0 — 100 km
Coverage rates, stock status, and cold-chain alerts plotted across every facility in your programme. Filter by country, vaccine type, and time period — and drill from country to clinic in a few clicks.
11
Country programmes
1,240
Facilities live
08
Active alerts
Kano State HQ
Coverage 87.4% · 12 facilities
● 1 cold chain alert
VaxAI Vision · Field Dispatch
Issue 24 · May 2026
Why we exist
The story of last-mile immunisation is the story of the people who carry, count, and care for vials — and the systems that fail them.
VaxAI Vision is the operating layer for the people who keep vaccines viable across the last mile — combining live computer vision, ensemble forecasting, and cold-chain telemetry behind a single, calm interface.
It is built for ministries, NGOs, and field supervisors who do not have time to learn another dashboard. It speaks DHIS2, mSupply, OpenLMIS, and FHIR — and works offline when the signal does not.1
Up to a quarter of vaccines lose potency before they reach a patient. Our job is to make that gap visible — and then close it, vial by vial.
Fig. 1 — A cold box arrives at a rural health post in Kisumu County. Without continuous monitoring, up to 25% of vaccines lose potency before they are administered.
2.4M doses
kept viable across the last mile since the platform first deployed in partner facilities in 2023.
1 See docs / integrations for the full list of supported standards and offline-sync behaviour.
A fully-loaded demo with sample data from three partner programmes — no login, no sign-up. Walk through the dashboard, the AR scanner, and the forecasting engine in one session.
Programme
Programme overview
Doses tracked
2.41M
+18.4%
Stock value
$1.42M
+6.1%
Chain uptime
98.7%
+0.6pp
Active alerts
8
−2
Stockout risk · 12-week forecast
PROPHET + LIGHTGBM
Aggregated across active deployments since 2023. Programme-level dashboards with country breakdowns available to partners and donors on request.
2.41M
Doses tracked
+18.4% YoY
1,240
Facilities live
Across 11 countries
98.7%
Cold-chain uptime
+0.6pp vs baseline
$4.1M
Wastage avoided
Modelled across pilots
“
For the first time, I know what is in our cold box before I open it.
A. Okafor
Senior health worker, Kano State
“
Forecasts that finally match what we see in the field. Our wastage dropped twelve points in the first quarter.
Dr. M. Wanjiru
Programme lead, Nairobi
“
The DHIS2 sync was live in a day. We spent the rest of the pilot actually using the data.
S. Nkosi
M&E, regional ministry of health
Most pilots start with 5–10 facilities and 90 days. Tell us what you're trying to fix — we'll come back with a proposed scope within a week.
AI-driven vaccine supply chain intelligence for last-mile immunisation programmes.
254 Chapman Rd, Ste 208
Newark, Delaware, 19702