Amine Ferdjaoui
PhD student in Data science and project leader at SogetiLabs, Capgemini.
Paris, France
About me
My journey:
Passionate about machine learning, my academic career began with a master's degree in machine learning, supplemented by a second one in MIAGE for project managment, and two years practical experience as a data scientist and full stack developer. I decided to explore the depths of the field further by starting a PhD with SogetiLabs and the Borelli Center. My research focuses on NLP, in particular the inference of causal relationships from unstructured textual data.
This exciting career journey has led to numerous publications, both national and international, as well as the development of two NLP software applications. I also share my knowledge by teaching master's students at the Université Paris Cité. Using the CIFRE (Convention Industrielle de Formation par la Recherche), I actively bridge the gap between theoretical exploration and practical application, contributing to real-life, industrial projects within SogetiLabs.
Finally, some quick bits about me.
- Double master's degree : machine learning and project managment
- Project leader
- 5 years+ experiment in innovation and web development
- University teacher
Experience
PhD & Project Leader - Capgemini
- Responsible for internal projects at SogetiLabs.
- Operational management and monitoring of development teams.
- Development of clustering and recommendation models using NLP techniques.
- Definition of requirements and discussions with customers to produce mockups.
- Company tutor for work-study students and interns.
Jul 2022 - Present
Data Scientist, Orange - Enovacom
- Development of a search and visualization platform for unstructured data in a healthcare data warehouse.
- Collecting, preprocessing and classifying Harvard Medical School's health data n2c2.
Apr 2021 - Oct 2021
Data Scientist & Full Stack developer, Numendo
- Maintenance and retraining of text classification models.
- Development of an rtsp video streaming platform (React and Flask).
Oct 2020 - Jan 2021
Data Scientist & Full Stack developer, Orange
- Creation of incident ticket classification models for Orange Cloud for Business.
- Realization of a model and a platform (Front and Backend) for detecting the wearing of masks in real time.
- Designed and managed a project to re-train and update models in real time.
- Creation of a data labeling platform for training AI models.
- Deployment and production launch of models in Docker on the Orange Flexible Engine Cloud.
- Developed the AI Marketplace api in php-Graphql.
Sep 2019 - Oct 2020
Data Engineer, Saynova
- Data analysis and visualization with Elasticsearch and Kibana.
- Log file collection and transformation with Logstash and Beats
Mar 2019 - Jun 2019
Publications
Talks
Meeting of the French-Speaking Classification Society 2024
International Mathematical Meeting Centre, Marseille, France
Presentation of a new explainable clustering algorithm applied on large textual data at meetings of French Classification Society.
ACM International Conference on Web Search and Data Mining 2024
Mérida, Mexico
Demonstration and presentation of the article "WordGraph: a python package for reconstructing interactive causal graphical models from text data".
French-speaking Conference on Knowledge Extraction and Management 2024
Dijon, France
Presentation and demonstration of the article "Modèles graphiques causaux interactifs pour les données textuelles" and the web application WordGraph.
AI-DSCY: Machine Learning Wordkshop 2023
Paris, France
Presentation of our recent work on exploration of causal graphical models applied on textual data using Multivariate Information-based Inductive Causation and co-clustering.
French-speaking Conference on Knowledge Extraction and Management 2023
Lyon, France
Presentation and demonstration of CORPEX: an exploratory web application of large textual data using robust co-clustering.
IPOL Journal: Image Processing On Line 2023
Paris, France
Presentation and demonstration of a new deep auto encoder consensus approach for clustering called "CAEclust: A Consensus of Autoencoders Representations for Clustering” (2022). In: Paris, France: Image Processing On Line, pp. 590–603.
Get in touch
Feel free to reach out to me if you are looking for a data scientist, an advice, or simply want to connect.
ferdjaouiamine@gmail.com
You may also find me on these platforms!