Analytics
The HIC analytics stack is the data layer that transforms raw health data from nine source systems into trusted, queryable analytics. It handles ingestion scheduling, multi-layer SQL transformation, distributed processing, and BI consumption.
dbtSQL-based data transformation across 4 model layers (staging → DWH → reporting). Powers all Superset dashboards.PrefectWorkflow orchestration for data ingestion from 10+ health source systems, scheduled reporting, and dbt runs.SparkDistributed processing for large-scale analytical workloads, accessible via Livy REST API.Spatial Analysis & GISGeospatial mapping of health and climate indicators onto Rwanda's administrative geography with GeoPandas.
Consuming tools
Transformed data in the warehouse is consumed by two downstream tools:
- Apache Superset (
superset.nhic.moh.gov.rw) — dashboards and ad-hoc SQL queries for health program reporting. Superset connects directly to thereportinglayer views produced by dbt. - JupyterHub — interactive Python notebooks for data scientists. Connects directly to the data warehouse for exploratory analysis, model development, and custom extracts.
Last updated on