Experience

2017 - Now : PhD Student and Data Scientist at TabMo (France)

Context : TabMo is an adtech company running Hawk, its own and operated mobile programmatic full stack platform. We have to deal with millions of bid request every second. Many statistical challenges arose from this field, such as displaying an advertisement at the right place and the right time. A crucial aim for adTech company is to be able to predict the probability of click in real-time among the voluminous and heterogeneous data.

Objectives :

• Clustering of mobile advertising campaign : developing a clustering method of publicity campaign time-series, i.e. of non-Gaussian longitudinal data, with a mixture of generalized linear models
• Predictive model in real time giving a probability of click

2015 - 2017 : Data Analyst in R&D Team at Acelys Informatique (France)

Analysis of urban data such as bicycle station filling data or others public infrastructures for transportation. The main objective was to develop a predictive algorithm able to inform about the number of available spot/bike for each bike station and the same for car park in real-time.

Internship :

• 2015 (6 months) : Junior Data Scientist at IBM Client Center of Montpellier (France)
• 2014 (3 months) : Statistician at Météo France in Toulouse (France)