17.8292° S, 31.0522° E · HARARE, ZIMBABWE

Rutendo Sibanda

GIS Developer / Geospatial Analyst

Building geospatial data infrastructure for climate and health research across sub-Saharan Africa.

Currently contributing geospatial exposure proxies to the CHEAQI–MNCH air-quality protocol and keeping the PALS research server online

[email protected] · Based in Harare, Zimbabwe

Countries across my project network

Participants characterized (PRECISE + HE²AT heat data)

Communities mapped across sub-Saharan Africa

Variables across 6 domains — heat, air pollution, climate, land cover, wealth, accessibility

Field Network

Direct fieldwork through the PRECISE study across Kenya, Mozambique and The Gambia, plus geospatial exposure models (population-weighted road-network density, road-proximity, fire-smoke, land-surface temperature) that I've built and that now feed research across the HE²AT Center / DS-I Africa network's 33-country footprint in sub-Saharan Africa.

hover to read coordinates

Direct fieldwork & data HE²AT / DS-I Africa / eLwazi consortium hubs Africa-wide tools I've built HE²AT / DS-I Africa network span (33 countries)

About

Geomatics engineer working as a Junior GIS Developer at CeSHHAR Zimbabwe, within the NIH-funded HE²AT Center consortium and Place Alert Labs.

I build the geospatial and AI-assisted infrastructure that climate-health researchers across sub-Saharan Africa use to link environmental exposure data to health outcomes. Skilled in geodatabase design and validation, spatial data quality assurance, Python-based geospatial pipelines, and translating spatial analysis into maps, dashboards, and tools that non-specialist stakeholders can use directly. Experienced in server administration for shared research platforms and in communicating geospatial findings to donors, researchers, and community partners. I work daily inside transdisciplinary teams — climate scientists, epidemiologists, data engineers, clinicians and community fieldworkers — translating between their tools and vocabularies so a single validated dataset can serve all of them, across time zones, power outages, and patchy connectivity.

Education: BSc Honours, Geomatics Engineering (Geoinformatics), University of Zimbabwe, 2020–2024, Upper Second Class. Final-year project: a population-weighted GIS model for spatially explicit assessments of healthcare access and traffic-pollution exposure, using population rasters rather than village centroids so results reflect where people actually live — later extended and published (see Publications).

Experience

Junior GIS Developer Feb 2025 – Present

CeSHHAR Zimbabwe, HE²AT Center / Place Alert Labs

Promoted from Geospatial Science Intern in recognition of performance on the PRECISE study.
Led the geolocation module for HarmonAIze, an AI-assisted multi-source geocoding and validation platform, applied to 451 locations across 21 studies in 10 African countries with a persistent validated-geocoding cache (see Projects).
Built exposure–outcome pipelines for the PRECISE study, linking multi-trimester temperature, PM2.5, and fire-smoke exposure to birth outcomes, including a natural-language query interface over DuckDB.
Scaled exposure assignment past 100,000 pregnant women and 40+ variables by designing a location-date deduplication method, then built PALs Earth, a self-service extraction pipeline now being adopted by a senior research fellow at the PRECISE Network (King's College London) for the PRECISE and DYAD datasets.
Contributed population-weighted road-network density and road-proximity exposure models as validated air-pollution proxies for the CHEAQI–MNCH research protocol (Journal of Global Health Economics and Policy, 2026), extended Africa-wide via an API and application.
Built and administer PALSHub (shared research cloud interface) and the PALSlab JupyterHub research environment; lead PALS Lab server infrastructure, including diagnosing a memory failure from abandoned kernels consuming 80+ GB of RAM, a domain migration, and a Cloudflare split-hosting design for outage resilience.
Developed an AI-powered research catalogue querying live participant data across 6,687 participants in three countries, now a core consortium communication tool, and built a WhatsApp-based environmental early-warning prototype (Open-Meteo, Twilio).
Designed and delivered a GeoAI workshop for the Zimbabwe Young Surveyors Network; co-developed a self-paced GeoAI-in-health course now serving 40 students; conduct qualitative thematic analysis of multilingual focus-group transcripts on climate, health and health-workforce resilience in Zimbabwe.
Co-authored a peer-reviewed paper measuring spatial access to schools and health facilities in Zimbabwe, published in the Statistical Journal of the IAOS (2025); containerised the workflow behind it and represented the HE²AT Center at the DS-I Africa Harmonisation Weeks in Cape Town (2025, 2026).

Geospatial Science Intern Jan 2024 – Feb 2025

CeSHHAR Zimbabwe, PRECISE Study

Assigned to create and lead development of accessibility metrics (access to care, road-network density) in ArcGIS; built the access-to-care model on population rasters with a population-weighted average, which became my dissertation and then a co-authored publication.
Geocoded pregnancy-cohort locations across 500+ neighbourhoods in The Gambia, Kenya and Mozambique; moved from manual searching to AI batch processing, identified accuracy failures (coordinates falling outside the target country), and used that experience to design the AI-assisted pipeline that became the HarmonAIze geolocation module.
Built a road-quality index and accessibility model for the study areas, and contributed to the PRECISE study on environmental determinants of maternal health in Kenya, Mozambique and The Gambia (6,687 participants, 525 communities), using H3 hexagonal tessellation for road-network density and accessibility models.
Developed Google Earth Engine and Python pipelines pulling ERA5, MODIS LST, MERRA-2, GHAP PM2.5, CHIRPS, NDVI, and soil/urbanisation layers.
Delivered geospatial training and technical support to interdisciplinary research teams across five countries.

GIS Consultant Intern Apr 2023 – Dec 2023

Technical Systems Zimbabwe

Contributed to Zimbabwe's national Ecadastre land information portal: automated digitisation and capture of physical survey records using Python (pytesseract, Pillow), map digitisation, and shapefile-to-image conversion.
Delivered client-facing GIS mapping and data management outputs across concurrent deliverables in a fast-paced consultancy environment.

Featured Projects

DS-I Africa Hub, Consortium

HE²AT Center

The Heat and Health African Transdisciplinary Center develops data science approaches to the health impacts of climate change: from heat-exposure modelling in pregnancy to city-level early warning systems in Abidjan and Johannesburg. My work on HarmonAIze and the PRECISE exposure pipelines feeds into this wider data ecosystem.

Climate & HealthConsortiumData Ecosystem

Published protocol · models contributed, not an author

CHEAQI–MNCH Air Quality Protocol

My population-weighted road-network density and road-proximity models are used as validated static proxies for vehicular-related air pollution exposure in this multi-country protocol linking air quality to maternal, newborn and child health outcomes across sub-Saharan Africa. I contributed the geospatial exposure models used in the study design; I am not a listed author on the paper.

Air QualityRoad-Network ProxyMNCH

doi.org/10.7189/001c.155593 →

Open-source, multi-country research consortium

HarmonAIze / Geolocation Module

HarmonAIze is an open-source toolkit that harmonises and integrates climate and health data for research teams in resource-constrained settings, using AI-assisted workflows and federated approaches that preserve privacy. I lead its geolocation module: it queries HDX, ArcGIS, OpenStreetMap and Google Maps in parallel, uses AI to parse free-text place names, and validates coordinates by proximity and reasoning, with maps for fast human review. Tested on 451 locations across 21 studies in 10 African countries — inter-source disagreements above 5 km affected 59% of locations, and a persistent validated cache now serves 93% of locations without external API calls, cutting mean processing time from 2.69 s to under 0.1 s per location.

PythonGeocodingData Validation

Live tool · PRECISE Network (KCL) adoption in progress

PALs Earth

A self-service web platform letting researchers upload participant locations and dates and extract harmonised environmental exposure data from Google Earth Engine — no Earth Engine scripting required. Built after exposure assignment needed to scale past 100,000 pregnant women and 40+ variables across 2000–2023, which required a location-date deduplication method so shared exposures aren't recomputed per person. A senior research fellow at the PRECISE Network (King's College London) is adopting it for the PRECISE and DYAD datasets.

Google Earth EngineSelf-Service PipelineScale

Kenya, Mozambique, The Gambia

PRECISE Study Exposure Pipelines

Built exposure–outcome analysis pipelines for the PRECISE Network's research into hypertension, fetal growth restriction, and stillbirth, linking temperature, PM2.5, and fire-smoke exposure data to birth outcomes. I also built the PRECISE Catalogue itself, including the GeoAI chat interface that lets researchers query the participant dataset in plain language.

DuckDBClimate DataHealth Outcomes

Prototype

WhatsApp Environmental Early Warning

Built using open weather data and Twilio to deliver climate-health alerts over a mobile channel usable by non-technical recipients.

TwilioOpen Weather DataEarly Warning

Zimbabwe Young Surveyors Network

GeoAI Workshop & Course

Designed and delivered a GeoAI workshop; co-developed an open-source, self-paced GeoAI-in-health course now serving 40 students.

TeachingGeoAIOpen Curriculum

Collaborations

My work only exists at the intersection of disciplines — every project pulls together people who don't normally share a table.

None of the projects above are solo work. HarmonAIze and the PRECISE pipelines are built with climate scientists, obstetricians, epidemiologists and software engineers reading the same data differently and needing it to agree. The HE²AT Center is explicitly transdisciplinary by design, spanning data science, public health and urban planning across its DS-I Africa partner sites. Day to day, that means translating a climatologist's NetCDF file into something a clinician can query in plain English, and making sure a community fieldworker in Mozambique and a data engineer in Cape Town are looking at the same validated coordinate.

Consortia & Networks

HE²AT Center / DS-I Africa

Climate science · Epidemiology · Data engineering · Public health

Transdisciplinary by design: climate scientists, epidemiologists and data engineers building shared heat-health exposure models across the DS-I Africa partner network.

PRECISE Network

Obstetrics · Environmental epidemiology · GIS · Community fieldwork

Multi-country maternal health study across Kenya, Mozambique and The Gambia, pairing clinicians and environmental scientists with community field teams.

HarmonAIze Consortium

AI/ML engineering · Health informatics · Geospatial science

Open-source, federated toolkit built alongside software engineers and health data scientists across a growing multi-country study network.

eLwazi ODSP (UCT)

Bioinformatics · Cloud infrastructure · Open data science

Pan-African open data science platform — containerisation and cloud-infrastructure collaboration with teams across the continent.

Zimbabwe Young Surveyors Network

Geomatics · Education · Youth development

Cross-generational knowledge-sharing: designing and teaching a self-paced GeoAI-in-health curriculum for early-career surveyors.

Tools & Platforms

Beyond HarmonAIze and the PRECISE pipelines, I build and help maintain the applications behind the PALs portal.

Built & Lead

PALSearthPoint-and-extract environmental data, NDVI, land-surface temperature, rainfall, elevation, soil, for any location and date, powered by Google Earth Engine.
Environmental Exposure Catalogue35+ climate and environmental variables shaping maternal and newborn health across Sub-Saharan Africa, for the HE²AT Center.
Africa Road Network DensityContinent-wide H3 hexagon layer of road density from OpenStreetMap, a proxy for traffic-related air-pollution exposure — validated as a static air-quality proxy in the CHEAQI–MNCH protocol.
Africa Road ProximityDistance-to-nearest-road accessibility layer with population-weighted summaries, built for the whole continent — also used as a validated proxy in CHEAQI–MNCH.
PALSlab HubThe shared JupyterHub research environment for the PALS team: kernels, shared data directories, and direct database access.

Contribute & Maintain

GIPEXSatellite-derived exposure indicators, heat, NO₂, PM2.5, vegetation, night-lights, for study areas without ground monitors.
SPECTRATurns GPS tracks and wearable-sensor data into minute-by-minute personal environmental exposure profiles.

Skills

GIS & Remote Sensing

QGIS, Esri ecosystem, Google Earth Engine, ERA5-Land climate data extraction, geodatabase design

Geospatial Development

Python spatial libraries, PostgreSQL/PostGIS, DuckDB, Django/Celery, Docker

Data Quality

Multi-source geocoding validation, spatial data cleaning & QA/QC, metadata standards

Data Science

Exploratory data analysis, supervised & unsupervised ML, deep learning fundamentals

Web GIS & Communication

Interactive dashboards, natural-language query interfaces, technical documentation, workshop delivery

Infrastructure

Linux server administration (JupyterHub), cloud/static hosting design, Git-based deployment

AI-Assisted Development

LLM-based coding, data pipelines and documentation workflows, natural-language query interfaces over research datasets

Field Data Collection

KoboCollect, Survey123, GNSS, total station & levelling (undergraduate training)

Languages

English (fluent), Shona (native)

Publications & Recognition

Measuring access to schools and health facilities in Zimbabwe: A distance-based analysis of service islands

Watambwa, P., Sibanda, R., Chinwadzimba, Z., & Jakachira, R., in Statistical Journal of the IAOS, 41(4), 2025

A spatial accessibility framework scoring access to 1,654 health facilities and 8,000+ schools across all 64 districts of Zimbabwe, built on the H3 accessibility modelling I first developed as a final-year project. The methodology has been recommended for adoption in national frameworks.

journals.sagepub.com/doi/10.1177/18747655251393052 →

A Multi-Source Geocoding and Validation Framework for Spatially Precise Health Research in Sub-Saharan Africa

Accepted for presentation, AARSE 2026 — 30th Anniversary Conference, Harare, October 2026

Applied to 451 locations from 21 health studies in 10 countries; multi-source consensus reduced median coordinate error from 1.00 km to 0.70 km. Describes the HarmonAIze geolocation module (see Projects).