About Me

I am a dedicated AI Engineer with specialized expertise in Machine Learning, Deep Learning, and Natural Language Processing. My professional focus centers on developing intelligent systems that address complex real-world challenges through innovative technological solutions.
Following completion of my scientific baccalaureate, I undertook rigorous preparatory studies (Math Sup/Spé) at Lycée Thiers (Marseille) before gaining admission to ENSIMAG, where I specialized in Computer Science and Data Science through the prestigious MSIAM double degree program.
With a comprehensive foundation in computer science and extensive practical experience, I focus on architecting and deploying scalable AI solutions, managing the complete development lifecycle from conceptualization through production implementation.
Recently, I have been focusing on creating LLM-based apps via Agentic AI.
Apart from my insatiable interest in new technologies, my other hobbies are the following :
Ambition never gets oldLouis XVIII
Experience
AI Research Scientist - National Institute of Informatics
April 2025 - Present
Led the development of a new meta-labeling framework aiming at enhancing LLMs forecasting capabilities for Time Series.
As part of the Prendinger's LAB, I studied the Marcos Lopez de Prado work and implemented specialized fine-tuning architectures for adapting SOTA models (such as Autoformers, Chronos, TimeLLM ...) to financial time series (cryptocurrencies & stocks).
Main tools & libraries : Git, Optuna, Pytorch, VSCode (Python), WeightWatcher
Link to the paper : Coming Soon.Data Scientist Intern - Hygie31
June 2024 - August 2024
Developed KPI visualization tools for optical pharmacy data at Hygie31. As part of the DINSI team, I built data pipelines automating collection, cleaning and features extraction process. I also created a specialized database for efficient information retrieval.
Main tools & libraries : Git, Informatica, Power BI, Pycharm (Python), Snowflake
AI Assistant Engineer - Centre National de la Recherche Scientifique
June 2023 - August 2023
Extended a Deep Learning architecture for protein structure comprehension. Within the LJK team, I integrated Laplacian motion vectors into the ScanNet model to find binding sites more precisely.
Main tools & libraries : Git, Pycharm (Python), Tensorflow
Projects
AI-Powered Mail Filtering System
TODO : AI filtering mail with LangChain
Technologies: Python, LanChain
View on GitHubSkills
Programming Languages
Python
SQL
Bash
ML Frameworks
PyTorch
Scikit-learn
Hugging Face
Optuna
Cloud & DevOps
Snowflake
SQL Server
Informatica
Tools & Platforms
Git
LangChain
Power BI