Greater Paris Metropolitan Region
I am a Senior ML Engineer at Meta in London, working remotely from Paris. I previously worked a Senior ML Engineer and Tech Lead at Wefox in Paris, conducting R&D on insurance digitalization and risk prevention. My first industry experience has been as Applied Scientist at Samsung AI Labs in Paris, conducting applied research into Machine Learning (ML), Computer Vision (CV) and Natural Language Processing (NLP). I have received a PhD in CV/ML at Paris-Saclay University (France), working on pedestrian detection and tracking in high density crowds. My PhD experience has formed my expertise around the CV/ML domains while improving my problem solving skills, my methodological approach to science and stimulating my continuous learning attitude. I have obtained my bachelor and master diplomas with honors in Computer Engineering at University of Pavia (Italy). Such period has been crucial for developing my skills in programming, agile software development, and algorithms.
Part of the new Wefox Innovation team located in Paris, responsible of applied research in ML. Our main focus is to design and develop solutions for client risk prevention, while also backing Wefox towards the goal of 100% insurance digitalization. Topics of interest: - Multilingual NLP on text and structured documents. - Risk assessment and forecasting with transformer models.
Applied research in several CV and NLP projects around multiple innovative products. Projects: - 3D lines detection for autonomous driving (non-linear 3D modeling of road, road line segmentation models) - Chatbot for assistance on Samsung products: automatic construction assistance knowledge graph, NLU of user queries to guide the conversation through the graph - Research study on exploiting transformer pertaining to provide strong baselines in zero-shot and few shot conditions. Publication as first co-author at ACL 2020: Pre-training Is (Almost) All You Need: An Application to Commonsense Reasoning (https://arxiv.org/abs/2004.14074) - Design and training of transformer model for topic segmentation of meetings (per-sentence classification of topic switch with context sentence window). - Person (re)-identification for meetings - Supervised internship on super-resolution of text on whiteboard images Participated to a study around quantum computing technology, which led Samsung Catalyst Fund to invest in IonQ company. Implemented and tested quantum circuits to report return on experience to IonQ.
Subject de thèse: [MOHICANS] Analyse de foules denses pour la modélisation, la planification et la sécurité. [MOHICANS] Towards Modeling High-Density Crowds for Assisting Planning and Safety. Group: SATIE (Systèmes et Applications des Technologies de l’Information et de l’Energie) - ENS Cachan Ecole Doctorale: STIC - Sciences et Technologies de l'Information et de la communication.
C/C++ laboratories and software projects assistant
Dense motion detection from a moving camera by multiview constraints