Université Lumière Lyon 2 — Lyon, France
Advanced training in machine learning, statistical modeling, big data processing, and cloud-based AI systems.
Université Lumière Lyon 2 — Lyon, France
Specialized in data analysis, machine learning foundations, and software development.
IFRI — Cotonou, Benin
Solid foundations in algorithms, software architecture, backend systems, and full-stack development.
Projects Completed
Success Rate
Years Experience
Building AI systems that solve real-world problems
SNCF Réseau
Designed an anomaly-detection pipeline (Denoising LSTM Autoencoder + Isolation Forest) on 500M+ time-series records, reducing computation time by 40%. Achieved an AUC of 0.80 and optimized the solution for Azure Databricks to enhance scalability and robustness.
EffetB
Developed a hybrid recommendation system combining knowledge graphs and NLP embeddings, reducing cold-start issues by 30%. Applied GraphSage, FastRP, and clustering methods, with 3D T-SNE visualizations for insights. Delivered a validated POC for product integration.
Alibora SARL
Built a call-center agent management system with performance tracking and audio integration. Enhanced a property-management mobile app and contributed to ETL pipelines via Laravel API.
Technologies I use to build intelligent systems
Python
R
TensorFlow
PyTorch
Keras
scikit-learn
Pandas
NumPy
OpenCV
YOLO
OpenAI
LangChain
Hugging Face
LangGraph
Google Cloud
Databricks
Docker
Kubernetes
Recent work in machine learning and artificial intelligence
Sharing insights on AI, machine learning, and software engineering
solation Forest is one of those machine learning algorithms that seems almost too simple, yet surprisingly powerful. In this post, we’ll unpack how it works, when you should (and shouldn’t) use it
In this post, we explore how to detect anomalies in sequential data using a deep learning-based LSTM Autoencoder, followed by KMeans clustering
