Hi! I'm
I love coding, gaming and lifting weights.
This description was written by a human. Definitely. Otherwise it would sound way fancier.
[Insert motivational closer]
Projects
My personal and academic projects.
COMPREVE
Django-based web application designed for analyzing Twitch chat messages. It provides tools for uploading, searching, filtering, and analyzing chat data.
Pink AI
Interface that allows users to experiment with language models by adjusting LLM parameters for a customized experience. Built with Streamlit, it enables fine-tuning settings like temperature, top-k, and top-p to control response behavior.
Toxic Comment Classification
DistilBERT, GRU, and LSTM models to detect and classify toxic comments in online discussions, with evaluation through error analysis and visualizations.
SMS Spam Detection
Multinomial-Gaussian-Bernoulli Naive Bayes, XGBoost, Random Forest, and Decision Tree models are used to distinguish between spam and legitimate messages, with methods to handle class imbalance and improve performance.
Fake News Detection
Logistic regression and naive bayes approaches to identify and flag potentially false news articles.
Lyrics Generator
Bi-LSTM and embedding models to predict and generate song lyrics based on an initial input sequence with categorical cross-entropy loss.
Sentiment analysis
Multiple machine learning models, including RNN, TFIDF, and MLP, are trained and compared for sentiment analysis on both films and tweets.
BTC Predictor
The XGBoost model is updated daily with fresh data for accurate bitcoin predictions, featuring data processing, model training, and visualization via a Streamlit dashboard.
Image Classification
Deep learning model using PyTorch to classify images through transfer learning. By leveraging pre-trained neural networks, it efficiently classifies 102 flower species with optimized training time and improved accuracy.
Experience
My academic and professional experience.
Data Scientist
Steerway, Nancy, France
Developing the data pipeline of a code assistant, including information retrieval, RAG, language model optimization via pruning and quantization, fine tuning, with data analysis, visualization and evaluation integrated.
March 2025 - Current
Professional Project
UR ReSO, Université de Montpellier Paul-Valéry, Montpellier, France
Development of a Django-based corpus exploration website related to online violence on Twitch.
December 2024 - March 2025
Artificial Intelligence & Machine Learning Team
DiaspUra, Paris, France
Design, development, and management of a conversational AI solution.
January 2024 - June 2024
Master 2, Linguistic Data Sciences
Grenoble Alpes University, France
Advanced studies in natural language processing (NLP), machine learning, and programming.
September 2024 - Current
Master 1, Language Technologies
University of Turin, Italy
September 2023 - July 2024
Resume
Click here to download my resume.
DownloadContact
Feel free to reach out for collaborations, questions, or just to say hello!