I am a Lead Researcher in the Discover (www.discover.com) Data Science Research team. My research focuses on leveraging machine learning techniques to extract signals from data. We work with large tabular datasets to predict risk behavior for financial use-cases. Our use-cases go from feature noise cleaning, feature selection in high dimensional regimes, Embedding and clustering very large datasets, explainability methods for both supervised and unsupervised models, and finally supervised learning model development. I’m also a part-time Statistics PhD student at the University of Cape Town (https://science.uct.ac.za/department-statistics) where my PhD projects revolve around statistical models, complexity sciences, and networks: I’m interested in new methods to build artificial graphs from data at scale with certain properties. A big chunk my work has consisted in approximating latent variable models to clustering generative models. Finally I am looking into methods to learn ensemble of orthogonal supervised models.