Expertise

Six areas, one goal: reliable data, used at the right level.

We work across the full life cycle of a health information system, from the initial assessment to ownership by national teams.

DHIS2 architecture & national rollouts

Designing and setting up DHIS2 platforms at national scale, from metadata configuration to server administration.

DHIS2 · Tracker · Maps/GIS · DataStore · Linux

Aggregate and Tracker configuration
National master facility registry
Native DHIS2 apps and analytics dashboards
Server administration, security and backups
National DHIS2 roadmaps (assessment, resources, budget)

Example · Guinea National DHIS2 rollout: monthly reporting from more than 500 health facilities, integrating the HIV, malaria and maternal and child health programmes.

Health system interoperability

Making data flow automatically between patient records, laboratories, logistics and surveillance systems, so nothing is entered twice.

OpenFn · HL7 FHIR · OpenHIE · HAPI FHIR · REST

Exchange architecture aligned with OpenHIE principles
OpenFn workflows and custom adaptors
Metadata harmonisation and semantic mapping
Continuous integration with automated tests
Technical documentation and knowledge transfer

Example · Burundi Connecting SIDAInfo, IBIPIMO and DHIS2: automated HIV reporting and viral load exchanges, signed off by the Ministry before going live.

One Health surveillance & early warning

Signal detection systems that connect human, animal and environmental health, in line with Africa CDC and WHO guidance.

EBS · CEBS · DHIS2 Tracker · One Health

Design of event-based and community-based surveillance
Verification, triage and risk-based escalation workflows
Signal definitions shared across sectors
Governance standards, metadata dictionaries and validation rules
Performance indicators and dashboards

Example · Guinea National One Health EBS Tracker for the PReSeS-AOC project (World Bank funding), covering four signal streams: human, animal, environmental and security.

AI for public health

Predictive models that are useful in the field: validated over time, explainable, and built into the tools teams already use.

Python · LightGBM · SHAP · Power BI

Data pipelines combining DHIS2, climate and demographic data
Prediction models with walk-forward temporal validation
An explanation for every alert, for district health officers
Web dashboards and native DHIS2 apps

Example · ANSS Guinea Predicting epidemic threshold exceedances (AUC 0.902): 58.6% of true alerts detected seven days ahead, across 19 diseases and 38 districts.

Application assessment & validation

An independent look at an existing tool: what works, what is missing, and what to prioritise so it can be used for the long term.

Assessment · Requirements · UAT

Assessment of functionality, usability and data flows
User needs gathered at national, regional and facility level
Prioritised requirements and gap analysis
User acceptance testing through to formal sign-off
Rapid supply chain assessments

Example · Morocco Assessment and validation of a national epidemiological surveillance application for FHI 360 (STRIDES), including digitisation of the risk assessment tool.

Training & web platforms

Self-sufficient teams, and knowledge platforms that stay online and up to date.

Training · SOPs · Drupal · Pantheon

Training for users and trainers
Standard operating procedures and user guides
Monitoring and evaluation portals powered by DHIS2
Maintenance and support for institutional websites

Example · Kenya, Tanzania, Mozambique SCDataUse portal: Drupal-based monitoring and evaluation dashboards integrating DHIS2 data from three countries.

How we work

Step 1 Assessment Understand the current system, the stakeholders and the needs at every level.
Step 2 Design Functional specifications validated with stakeholders.
Step 3 Build Configuration, integration and automated tests.
Step 4 Validation User acceptance testing through to formal approval.
Step 5 Handover Deployment, training and documentation so teams can run it themselves.