Case study · 2026

Maternal Health Analytics

Well on Their Way — turning fragmented Excel health databanks into analysis-ready tables, GIS layers, and Tableau dashboards for maternal–child programs in Gulu District, Uganda.

PythonPandasTableauQGISPlotlyGeoPandas

Overview

Well on Their Way supports maternal and child health through village health team (VHT) training and facility-level monitoring. Program data lived in large Excel “databanks” that were hard to refresh and join to maps. I built a repeatable Python pipeline, cleaned spatial boundaries in QGIS, and published interactive Tableau dashboards for training coverage, district impact (Gulu vs Omoro), and obstetric POCUS field data.

Data pipeline

Excel sheets were imported into long-format CSVs, then reshaped into analysis tables with population denominators and inferred coverage rates for Impact and VHT Training metrics. Plotly scripts produced HTML chart hubs for quick exploration before dashboarding.

Plotly chart from the cleaned Impact or VHT training pipeline
Pipeline output — interactive Plotly charts from cleaned long-format data

GIS & QGIS

Public and local shapefiles needed cleanup before joining metrics. Boundaries were fixed in QGIS, exported as GeoJSON, and joined to VHT and impact indicators so Tableau maps aligned with real subcounties and health-center catchments.

QGIS map of cleaned Gulu subcounty boundary layers
QGIS — cleaned and aligned district / subcounty boundaries

Tableau dashboards

Dashboards and storyboards covered antenatal and postnatal coverage, immunization and nutrition indicators, Gulu vs Omoro impact comparisons, and POCUS survey results — filtering by period and geography on the cleaned layers.

Interactive demo

Live Tableau embeds are withheld while program data stays private. An interactive walkthrough is available on request; a scrubbed Tableau Public link can be added here later.