Case study 04
Amakomaya health platform
An award-winning maternal and neonatal platform, and the interoperability architecture underneath it.
Problem
Antenatal and postnatal care depends on information reaching women at the right point in a pregnancy — and that is exactly where rural and underserved populations are least well served. The gap is not usually clinical knowledge. It is distribution.
Building for that context imposes real constraints: intermittent connectivity, low-cost devices, and a requirement that whatever gets collected can still exchange cleanly with national systems rather than becoming another data island.
Approach
I co-founded Amakomaya and lead its technology function — architecture, AI adoption, engineering team, and the production stack.
- Architected an AI-enabled platform combining clinical decision support, patient education content, and standards-based health data exchange.
- Designed the interoperability layer on OpenHIE architecture, connecting point-of-service systems, national registries, and DHIS2.
- Implemented HL7 FHIR resources and profiles, SNOMED CT mapping, and ICD-11 terminology services so the data is semantically interoperable.
- Established the full production stack — Docker, PostgreSQL, Supabase, Keycloak for identity and access, Nginx, hardened Linux — with GitHub-based CI/CD.
- Lead multidisciplinary Agile engineering teams across the complete SDLC, through DevOps delivery, monitoring, and user support.
Technology
- OpenHIE
- HL7 FHIR
- SNOMED CT
- ICD-11
- Docker
- PostgreSQL
- Supabase
- Keycloak
- Nginx
- Linux
- CI/CD
Impact
Amakomaya has been recognised nationally for advancing maternal and neonatal health education and access among rural and underserved populations, and its implementations are built on open-source components that others in the digital health community can reuse.
- National recognition for advancing maternal and neonatal health education and access.
- Standards-based exchange with national registries and DHIS2 rather than an isolated data store.
- A secure, reproducible, cost-efficient deployment model.
- Open-source stewardship — DHIS2, OpenMRS, OpenIMIS, and Open Concept Lab implementations reusable by the wider community.