/ Mario Souto

I'm an engineer specializing in AI, control, and optimization, with a focus on enhancing energy systems and battery storage. After earning my PhD in Electrical Engineering, I worked at AWS, where I developed AI models to optimize Amazon's renewable energy portfolio. I was the tech lead for the Renewable Energy Optimization (REO) team, and the technology I developed contributed to Amazon becoming one of the world's largest corporate buyers of renewable energy. Later, I was the tech lead at Equilibrium Energy, where I developed and deployed grid-scale battery control and optimization algorithms.

During my PhD, I created a novel algorithm and solver for Semidefinite Programming problems, a somewhat exotic subset of convex optimization problems. Since then, my research has revolved around model predictive control, optimization algorithms, and machine learning.

Contact

LinkedIn - Twitter - Github - Google Scholar

<my-first-name> h <my-last-name> at gmail dot com

Open Source

ProxSDP : Semidefinite Programming (SDP) optimization solver.
StateSpaceModels.jl : Julia package for modeling, forecasting, and simulating time series in a state-space framework.

Research Papers

Souto, Mario, Joaquim D. Garcia, and Álvaro Veiga. “Exploiting low-rank structure in semidefinite programming by approximate operator splitting.” Optimization.
Souto, Mario, et al. “A high-dimensional VARX model to simulate monthly renewable energy supply.” 2014 Power Systems Computation Conference. IEEE, 2014.
Von der Weid, J. P., Souto, M. H., Garcia, J. D., & Amaral, G. C. “Adaptive filter for automatic identification of multiple faults in a noisy OTDR profile.” Journal of Lightwave Technology.
Souto, Mario, Joaquim D. Garcia, and Gustavo C. Amaral. “ℓ1 Adaptive trend filter via fast coordinate descent.” 2016 IEEE Sensor Array And Multichannel Signal Processing Workshop
Saavedra, Raphael, Guilherme Bodin, and Mario Souto. “StateSpaceModels. jl: A julia package for time-series analysis in a state-space framework.” arXiv preprint.