"I would describe myself as a very motivated person, towards both my job and my life as a whole. I am very passionate about transforming numbers into knowledge and using them for optimal decisions, while always thinking out of the box.\n\nI speak:\n- R, Matlab, Python, Spark, SQL, CPLEX&GAMS.\n- Machine learning and data analytics, data visualization\n- Constrained and non-linear optimization, robust and stochastic programming, MPC, paralelization and large-scale decomposition\n\nOn a personal level, I consider myself to be a very positive person, with an easy smile and full of energy. I spend a couple of hours per week doing sport, mostly rock climbing. From climbing I learned how to focus, and how important is to communicate well with your climbing team. Also, two years ago I cycled unsupported across South America: I love exploring new places at a slower pace, preferably in close contact with nature and with local people. From traveling, I have become more open minded and learned how to adapt to any situation."
My main tasks are twofold. Firstly, to assign warranties on the performance of wind turbines and quantify the associated risk. Secondly, to analyze data from measurement campaigns and detect possible misalignments on the promised performance. I am also responsible for developing R-shiny tools and to maintain our team database.
Some of the daily challenges are first to extract meaningful information from multi-dimensional datasets and then communicate it in a simple manner to electrical engineers and sales officers. The title of my PhD is "Stochastic Energy Systems". My research interests are optimal decision making under uncertainty, new forecasting methods and machine learning techniques. More specifically, it focuses on stochastic programming and inverse optimization applied to smart grids. The research can be divided in two main topics.1. Smart grid. In (J. Saez-Gallego et al., 2016), we used an inverse optimization framework, which recast as a bi-level problem, to obtain a bid that represents best the response of the pool of consumers.2. Ancillary services. A probabilistic approach to the problem of determining the number of electrical reserves is published in (J. Saez-Gallego et al., 2014). We developed a stochastic optimization framework that combines statistical forecasts of demand, wind power production and outages in power plants. Two disciplines are mixed in my master degree. First, forecasting methods and advanced data analysis methods. Secondly, operations research techniques like stochastic optimization and decomposition techniques. I often use programming languages such as R, Matlab and GAMS. The main topic was statistical models such as time series, quality control and probability theory. The secondary topic was operations research, like networks and integer programming. The third topic was programming languages such as SQL, C++ and Java.