HealthTrack AI
AI-powered chronic disease management platform
12K+
Patients monitored
40
Clinics onboarded
94%
Early alert accuracy
−31%
Hospital readmissions
The Challenge
A German health-tech startup needed a GDPR-compliant platform that could ingest wearable vitals data and flag deterioration before hospitalisation. Existing tools were reactive, not predictive.
The Solution
We built a Flutter patient app + Python ML pipeline that ingests Bluetooth vitals, trains per-patient anomaly models, and alerts clinical staff via a web dashboard. The system integrates with existing EHR via HL7 FHIR.
How We Built It
Requirements
GDPR/HIPAA compliance mapping, EHR integration spec, ML data pipeline design.
Data Engineering
FHIR data ingestion, vitals normalisation pipeline, labelled training dataset creation.
Model Development
Per-patient anomaly detection models with explainability (SHAP values).
App & Dashboard
Flutter patient app with BLE device pairing; React clinical dashboard with alert management.
Compliance & Launch
MDR compliance review, penetration test, phased rollout across partner clinics.
"The ML pipeline Ubikon built is remarkably accurate. We saw a 31% reduction in readmissions in our first 6-month trial — that is a result we could not have achieved without their AI expertise."
Dr. Markus Bauer
CTO, HealthTrack AI
Tech Stack
Start a similar project
Free proposal within 24 hours. Fixed-price contract. 100% source code ownership.