Mael Laschon enhanced the HiggsML Notebook Starting Kit in the blackSwanCS/Higgs_collaboration_B repository, focusing on improving onboarding, reproducibility, and experimental workflows. Over the course of a month, Mael refactored the Jupyter Notebook structure, streamlined data loading, and reworked model initialization to ensure stable, repeatable experiments. The project migrated from a BDT to a neural network approach, aiming to improve predictive performance. Mael also introduced enhanced logging and integrated image-based visualization assets, supporting clearer result interpretation. By establishing cross-notebook analysis pipelines and leveraging Python, data cleaning, and version control, Mael delivered a more robust and user-friendly experimentation environment.