Client : Air France - Operations Research Service (From January 2016 to April 2016 - Following the acquistion of Bluestone Consulting by EY)Pursuit of previous projects, in addition :Project description : Building of a web application for the analysis of traffic dataTechnology : R ; R ShinyTasks : - Data management (Production of indicators ; aggregation) - Data Visualization

Client : Air France - Operations Research Service (From October 2013 to January 2016)Project description : Building of recommender systems for destinations and paid optionsTechnology : R ; Apache Pig ; R ShinyTasks : - Bibliographic research - Data management (Production of indicators ; aggregation). - Building of predictive modelings and recommender systems - Building of web applications for an interactive presentation of the POCs' results and data explorationProject description : Estimation of a market share modeling to simulate the impact of a modification of Air France's flight scheduleTechnology : R ; Apache PigTasks : - Bibliographic research - Data management (Production of indicators ; aggregation) - Estimation of the modeling - Building of web applications for an interactive presentation of the modelingsProject description : Study of the revenue dilution linked to the sale of Paid UpgradesTechnology : RTasks : - Data management - Production of indicatorsProject description : Challenging of existing churn detection modeling on the Flying Blue frequent traveller programTechnology : R on HadoopTasks : - Bibliographic research - Bootstrapped evaluation of predictive methods (Logistic regression ; Naive Bayes Classifier ; decision trees ; random forests). Bootstrapped evaluation has been parallelized on a Hadoop clusterClient : GDF Suez (February to September 2013)Project description : Building of a marketing clustering of Small and Medium Businesses MarketTechnology : SASTasks : - Data Management (Production of indicators ; aggregation ; imputation of missing values) - Principal Component Analysis - Hierarchical Ascendant Classification - Description of final clustersInternal projects :Academical project tutoring at Ensai (from January to May 2015)Subject : Delay prediction in the air trafficEnd of Studies internship tutoring (from April to October 2015)Subject : Topic detection in text data

Project description : Used statistical techniques to help better understand P&G’s brands’ performance in Western Europe on the Hair Care segmentTechnology : R ; Tibco SpotfireTasks : - Bibliographic research - Data management (Production of indicators ; integration of external data ; aggregation) - Principal Component Analysis - Building of a predictive modeling to predict brands' quarterly value sales

Researcher's assistant on a project involving the use of sampling techniques to build an efficient design and estimation pattern for multi-phase natural resource surveys