At SNCF Réseau, I led the migration of a strategic, multi-department BI performance tool from Excel and Power Query to Azure Databricks. The move cut data refresh time by about 60% and reduced monthly data preparation from about two days to three hours.
I designed and industrialized PySpark and Delta Lake pipelines on a Medallion architecture, with incremental loads, schema validation, and unit tests. I also built Power BI reports for national KPIs, including regional breakdowns and trajectories to 2028.
At Air Liquide, I optimized Python ETL processes and managed data pipelines from multiple sources. Earlier, at Saint-Gobain, I worked on Azure data and cloud projects, including a marketing data centralization solution and a FinOps project.

