Forecast sell-in using Machine Learning algorithms for automotive and pharmaceutical industries- Collected relevant and attainable data. Developed client specific features and Machine Learning model.- Implemented the model and proved its efficiency. Planned in-vivo tests with forecasters team.Master data quality management for pharmaceutical industry- Designed data quality testing procedures with operational teams within a complex Supply Chain environment. Implemented them in Dataiku to deliver a robust and evolutive tool.- Create PowerBi visual report to monitor data quality indicators and manage correcting actions.

Balancing cost and industrial flexibility in Supply Chain- Analysed ill-defined production planning and production multi-sourcing problems. Identified manufacturers' and consultants' needs to define the framework of future optimization methods.- Mathematically modelled industrial problems. Adapted risk measures concept from the financial industry to Inventory Management.- Developed new discrete and stochastic optimization methods applied to Supply Chain. Tested their efficiency on real data comparing heuristic and exact approaches.- Managed a long-term project respecting both industrial due dates and scientific rigor.- Presented my research results to students, consultants and international scientific committees.

Vehicle routing problem with time windows (VRPTW) for ERDF’s technicians - Analysed audit report of the organisation and optimisation of ERDF’s technicians. - Collected and analysed academic articles about VRPTW and adapted column generation algorithm for the ERDF specific problem. - Proposed the architecture of code and implement the algorithm. Tested its efficiency on real data and improved it using literature review and C++ specificity.

Computational complexity of balancing the stations of a self-service “bike hire” system on a circular graph- Collected and analysed academic articles. Synthesized the current knowledge on complexity of balancing graphs.- Proved theorems that give polynomial algorithms to optimize the balancing of stations in various specific circular graphs. Proved partial results and gave conjecture to help future research on the general case of circular graphs.

Automatic reduction of the degrees of freedom in a 3D-constrained system- Assimilated industrial programming processes to join a hundred-people-large development team. Took part in the development of state-of-the-art industrial software such as CATIA and SolidWorks.- Programmed back-testing C++ functions to check, save and replay operators’ performances. Designed and created C++ operators to control the diverse levels of auto-constraining run. Implemented the designed code within the official Dassault released code.- Took part in the department brainstorming committee. Suggested innovative and user-friendly interfaces which are now used in several departments of Dassault.- Prepared code documentation and user manual for internal communication (for internal clients and technologies department management) through reports and PowerPoint presentations.

Creation of a software to manage and optimize the planning of oral examination- Analysed an ill-defined planning problem. Interviewed clients to understand their needs and define the objectives, the input and the output of the application.- Created mathematical modelling adapted to an efficient solving and separated it in distinct easier problems. Developed and used heuristics and exact algorithms to solve them.- Defined code specifications. Implemented algorithms making use of existing solver (GLPK library).