Chief Data Officer · Data Scientist
Turning a live pulse of Kenya into better decisions.
I'm Charles Sweetland, Chief Data Officer at Stahili. We turn 40,000+ daily consumer survey responses into insight for businesses, policymakers and investors, using interactive visualisation, applied statistics and AI agents.
Nairobi, Kenya · London and Exeter, UK
01
Interactive data visualisation
Bespoke D3 graphics, from national dot-density maps to survey dashboards, built to explain rather than decorate.
02
AI agents & Claude
Agentic workflows, MCP servers and automated reporting pipelines that put LLMs to work on real operational data.
03
Applied statistics & ML
Causal inference, survey weighting and machine learning, from remote-sensing studies to wearable biomarkers.
04
Data platforms
Real-time field-data infrastructure capturing 40,000+ daily survey responses across Kenya.
Selected work
Case studies
Keeping the lights on: what 4,511 Kenyan small businesses told us about power cuts
A national survey with Boston Consulting Group on how unreliable grid power hits Kenya's micro, small and medium businesses: who stops trading, who stops investing, and what they'd pay to fix it.
- SME owners surveyed
- 4,511
- SME owners surveyed
- say outages cut or stop trading
- 79%
- say outages cut or stop trading
- would pay more for reliable power
- 74%
- would pay more for reliable power
A light in the dark: can satellites see mini-grid electrification?
Peer-reviewed research testing whether night-time satellite imagery, the basis of most electrification maps, can detect the 700+ mini-grids built across sub-Saharan Africa. Largely, it can't.
- mini-grids analysed
- 700+
- mini-grids analysed
- countries, 2014–2023
- 20
- countries, 2014–2023
- of known sites flagged as electrified by OEMaps
- 9%
- of known sites flagged as electrified by OEMaps
Which appliances actually grow mini-grid demand?
A multi-year field experiment across 25 mini-grid sites in Kenya, Nigeria and Tanzania testing whether appliance financing raises electricity consumption. Productive-use appliances did; household appliances didn't.
- consumption per user with productive-use financing
- +48%
- consumption per user with productive-use financing
- consumption of top-20% productive users vs peers
- 16×
- consumption of top-20% productive users vs peers
- connections across 25 sites
- 12,000+
- connections across 25 sites
About
A bit about me
As Chief Data Officer at Stahili, I lead our mission to make real-time, high-quality field data available to the organisations shaping Kenya's future. Every day, more than 40,000 people share survey responses through our platform, creating a live pulse of how the country thinks, feels and acts.
We use this data to help businesses, policymakers and investors make better decisions, whether that's improving access to energy, deepening financial inclusion or understanding shifts in consumer behaviour.
If you're building products, policies or partnerships that depend on understanding Kenya's people and markets, let's talk.
Experience
Oct 2025 – Present
Stahili
Chief Data Officer
Feb 2023 – Oct 2025
CrossBoundary
Head of Data Analytics
Aug 2021 – Feb 2023
CrossBoundary
Senior Data Scientist
Apr 2016 – Aug 2021
Springhead Data
Founder
Aug 2019 – Aug 2021
Vento
Chief Technology Officer
Oct 2015 – Aug 2021
Activinsights
Data Scientist
Publications & open source
Research, writing and code
Peer-reviewed research
- Journal articleMay 2026A light in the dark: detectability of mini-grid electrification with remote sensing in sub-Saharan Africa
A Difference-in-Differences model across 700+ mini-grids in 20 countries showing that night-time satellite imagery systematically underestimates rural electrification.
Isaiah Lyons-Galante, Charles Sweetland, Morteza Karimzadeh · Environmental Research: Infrastructure and Sustainability
- Journal articleDec 2025Associations between within-day step accumulation pattern and clinical measures of physical function
A change-for-change analysis of longitudinal data in community-dwelling older adults, validating the pfSTEP digital biomarker for physical function from accelerometer data.
Joshua Twaites, Charles Sweetland, REACT Research Team · International Journal of Behavioral Nutrition and Physical Activity
Articles & commentary
- CrossBoundaryJul 2025Second-life batteries offer up to 28% cost savings for mini-grid storage
Field testing of repurposed EV batteries across South Africa, Nigeria and DRC: 11–28% upfront savings when batteries last seven years or more. Cited on charge management and depth-of-discharge economics.
CrossBoundary Mini-Grid Innovation Lab · CrossBoundary Innovation Insight
- CrossBoundaryJan 2024Perspectives on 2024 — CrossBoundary Compass
Featured contribution on the expanding use of AI within CrossBoundary and the potential of generative AI in frontier-market investment.
CrossBoundary Leadership · CrossBoundary Compass
- NextBillionOct 2023What's the most efficient way to deploy mini-grid distribution infrastructure?
Preliminary results from a Mini-Grid Innovation Lab study comparing distribution-grid deployment approaches, with a clear winner for commercial sustainability.
Charles Sweetland, Tombo Banda, Faisal Olanipekun · NextBillion
- CrossBoundaryJan 2023Appliance financing for mini-grid profitability
Productive-use appliance financing lifted average consumption per user by 48% and mini-grid profitability by 11–44%, using regression, propensity-score matching and random forests.
CrossBoundary Mini-Grid Innovation Lab · CrossBoundary Innovation Insight
Open source
- R packageFeb 2024GENEAclassify
Segmentation and classification of GENEActiv accelerometer data. Used in clinical trials and 1,500+ peer-reviewed studies.
Charles Sweetland et al. · Activinsights · CRAN
- R package2016GENEAsphere
2D and 3D visualisation of raw or segmented accelerometer data, complementing GENEAclassify.
Charles Sweetland · Activinsights · CRAN
- R package2015GENEAread
Reads GENEActiv binary files into R: header extraction, downsampling, time intervals and auto-calibration.
Zhou Fang, Joss Langford, Charles Sweetland, Jia Ying Chua · Activinsights · CRAN
Education
2014 – 2015
MSc Advanced Mathematics
University of Exeter
2011 – 2014
BSc Mathematics
University of Exeter
2003 – 2011
A-Levels & GCSEs
Axe Valley Community College
Mathematics, Further Mathematics and Physics at A-Level