During my internship at LVMH, I worked closely with the network, operations, and cloud teams to implement two generative AI-based tools. These tools, now in production and used daily by 100 engineers, aim to improve the processing and resolution time of technical incidents across the 75 LVMH houses worldwide.The first application is a tool for identifying key information, rephrasing, and reformatting an incident ticket. It is used to create a structured email that will inform the concerned parties of the incident in progress. To do this, I fine-tuned Llama2 on a dataset that I built from hundreds of examples.The second application is a Chat that answers engineers' questions for which the answers are found in a vast database of internal LVMH IT documents. These documents may be old resolved incidents, nomenclatures, or documentation. This tool is based on RAG technology adapted to the technical documents of the use case, to which a keyword search is added.During this internship, I had the opportunity to work on all stages of the project, from defining the need to implementing it in a cloud environment.

During my internship at Erasmus MC laboratories, I had the opportunity to design, implement, and evaluate weakly supervised segmentation methods for medical images. My central role involved developing an innovative approach to annotate a set of images into distinct segments. Concurrently, I devised and integrated anatomical constraints to guide the training of a neural network.One of my major contributions was the creation of a boundary segmentation function between two distinct regions. This function enhanced the segmentation accuracy by precisely delineating anatomical boundaries between these regions. My involvement in this project enriched my skills in programming, image processing, and manipulation of neural networks for intricate tasks.This internship also provided me with the chance to collaborate closely with a team of experienced researchers and engage in constructive discussions about advancements in medical imaging. I take pride in having contributed to the exploration of novel approaches in the realm of medical segmentation, and I am eager to continue making contributions to scientific research at the crossroads of technology and medicine.

As Treasurer of the ENSTA Paris cultural association (500 members), I was responsible for managing the association's budget (60k€/year) and seeking grants to fund artistic projects (cultural weekend in Europe, art festival, concerts, etc...).I was also the driving force behind inclusion projects such as the International Week, which aimed to highlight the cultures of the school's international students.

As a driver for Transports CASSIER, I was responsible for daily collections of 600 used tires from automotive establishments in the Grand Ouest region of France. This activity is part of a recycling and recovery industry that handles five hundred thousand tons of tires per year.