Yassine

yassine.jpg

Meet Yassine! He’s an intern at Total E&P in France, working in the Strategy, Planning, Portfolio & Performance division hosted by the Data Analytics Competence Center (DACC)

 

“I work in the S3P division (Strategy, Planning, Portfolio & Performance) in Paris. During my internship, I am being hosted by the Data Analytics Competence Center (DACC) and I’m working on operations research and data science subjects.

 

One of the most exciting projects I have worked on is the development of a tool for optimizing the exploration portfolio. This proposes the optimal selection of petroleum prospects to maximize profit within set constraints, such as budget. By using this tool, we can optimize the selection, timing and the level of investment of each prospect.

 

I am currently an engineering student at Telecom Bretagne, having previously obtained a License (equivalent to a Bachelor’s degree) in Statistics and Stochastic Simulation at the University of Bretagne Occidentale, in parallel with my engineering course. After a gap year, I intend to focus my studies on Applied Mathematics, specializing in Data Science.

 

What I like most about my internship at Total is the opportunity to work on important subjects tackling real business problems – as an intern, I am considered a full member of the team. After some initial training, I was quickly assigned to projects with high added value for the business. I benefit from regular feedback from my team and learn every day from the very competent people around me.

 

I started to be interested in Data Science, without having previously had the opportunity to apply it to real business problems. Participating in the Total Data Challenge competition was a good opportunity for me to apply and enrich my knowledge. This is a competition organized by Total’s Group Data Officer and Chief Digital Officer and the goal is to promote a Data Driven culture within the group and identify people interested in this field. Nearly 250 employees of different levels of experience went head to head over the 3-month Challenge, to solve 3 problems in Marketing, Production and Refining.

 

To give an example, the Marketing challenge involved constructing a forecasting model to estimate the national demand for fuel oil in each French department over a 15-day horizon. A large part of the work was cleaning, visualizing and organizing the data in order to obtain the maximum insight, before we started modelling. This step requires some knowledge on the classical algorithms of machine learning and statistics.