About
I am a power sector specialist with technical roots in statistical signal processing, machine learning, graph theory, and optimization. I have extensive experience working at the intersection of energy and computation across startups, academia, and international organizations. I hold a bachelor’s degree in Computer Science and a Ph.D. in Electrical Engineering and Computer Sciences, both from the University of California, Berkeley.
My research has focused on developing algorithms and measurement methods for enhanced grid visibility to bolster renewable integration, system resilience, and operational efficiency. Designing algorithmic tools that succeed within real-world constraints—such as measurement noise, system uncertainties, sensor coverage and resolution limitations, and human user needs—has been a consistent theme in my work. I have also worked on a diversity of energy system modeling projects, from network simulation to capacity expansion planning.
Outside the energy sector, my love of signal processing has, in the past, led me to work in domains such as agricultural automation and medical imaging.
Interests
Grid measurement and visibility
Electric grids are measured at ever finer resolution, but turning noisy, incomplete measurements into a reliable picture of what the grid is doing is still hard. I work on algorithms and measurement methods for grid visibility, including topology and impedance estimation, event detection and clustering, and power quality monitoring, with a focus on high-resolution sensors such as PMUs.
Renewable integration and energy planning
Integrating renewables and storage, and extending reliable electricity to the people who still lack it, depends on good models and good data. I am interested in system optimization and capacity expansion planning, battery storage, and measuring power quality and reliability, including in Sub-Saharan Africa.
Signal processing and machine learning
I came to energy from statistical signal processing and still approach problems that way: models that are honest about noise, uncertainty, and limited sensing, and algorithms that people can understand and use. Past work includes compressed sensing, probabilistic streaming clustering, and learning from synchrophasor data.
Imaging and automation
Earlier in my career I worked on computer vision and medical imaging, from estimating sorghum biomass from robot-collected field images to simulating microscope images of neuronal activity.