Data scientist · Energy systems researcher
I use machine learning, geospatial analytics and optimization to plan electricity systems that reach people who don't yet have any.
My work sits between data science and power systems. I build spatial and statistical models that tell electrification programmes where to put infrastructure, forecasting models that anticipate how grids behave, and optimization models that size hybrid energy systems under real cost and reliability constraints.
Right now I analyse energy-access data covering six African countries and more than 48,000 mapped locations at the Access to Energy Institute, and research microgrid optimization in the Energy with AI Lab at Carnegie Mellon University. Research from that work has been accepted at IEEE Power Africa, IEEE APPEEC, IEEE ICAIGE, S4IOT26 and Deep Learning Indaba 2026.
I hold an M.S. in Electrical & Computer Engineering from Carnegie Mellon University and a B.Eng. in Electrical & Electronics Engineering from the Federal University of Technology, Minna.
Spatial models and benchmark datasets that help decide where minigrids and grid extensions should go, and in what order.
Mixed-integer dispatch and capacity-expansion models for hybrid systems serving agricultural, industrial and health-facility loads.
Time-series forecasting and anomaly detection on large-scale grid measurements, with the data pipelines that make them usable.
Scenario modelling of how energy transitions — electric vehicles, renewables, policy frameworks — change emissions and air quality.
Deep Learning Indaba 2026 · Top 3 Best Paper Poster
A high-resolution spatial benchmark dataset for machine learning in East African energy planning, built so that siting and prioritization models can be compared on common ground.
IEEE APPEEC 2026
A mixed-integer linear programming formulation for sizing and dispatching hybrid microgrids serving agricultural and industrial loads, balancing capital cost against unserved energy.
IEEE Power Africa 2026
A multi-stage capacity-expansion case study in Rwanda that models how clinic demand grows over time, so systems are staged rather than oversized on day one.
Ongoing · Access to Energy Institute
Spatial and statistical models of infrastructure readiness across six countries, delivered as interactive dashboards that replaced manual reporting cycles for country programme teams.
Carnegie Mellon University
Time-series forecasting and anomaly detection on large-scale grid measurement data, reaching demand forecasting error below 6% MAPE, with automated pipelines handling the preprocessing.
Applied machine learning, energy system modelling, intelligent systems.
Control systems, energy systems.
Open to research collaborations and roles in data science and energy systems — electrification planning, grid analytics, and optimization for energy access. Do reach out.