• Analyzed web user journeys to predict long-term purchase decision with RTB and Flaminem Data• Trained Logistic Regression models to score CRM clients and identify promising leads• Dealt with 40 T of data on 5 clusters with big data tools such as Scala, Spark, MR, Zeppelin and Hue

• Programmed classification and clustering data mining algorithms using R, then Scala and Git in order to provide a deep understanding of consumers’ behavior and detect future churn• Managed exploration of data from web navigation under Linux environment using Python and SQL.• Implemented a proximal stochastic gradient descent with a line search to fit a regularized logistic regression

• Used regression models to understand the effects of different advertising media• Developed collaboration between the various subsidiaries to improve data collection and sharing• Detected opportunities to use these data in a local advertising plan using Qlikview and exploit vast new flows of information