Summer 2025 · Marseille, France
Interned as a Data Science Intern at CMA CGM, a leading global shipping company, within the Security & Intelligence team. Focused on building anomaly-detection models, data pipelines, and full-stack applications to enhance maritime security and risk analysis.
What I worked on
- ETL pipelines in Dataiku and Snowflake, optimizing SQL performance to cut processing time of 800M+ container logs from ~20 hours to under 1 hour for near-real-time analysis of high-risk containers
- Full-stack features within an internal web application (React, FastAPI), building multiple frontend components and backend APIs used daily by 50+ analysts to streamline investigative workflows
- An algorithm using H3 spatial indexing to identify shippers' likely origin zones with 75%+ accuracy by filtering out hubs/ports and reconstructing average routes, for anomaly detection across global shipping patterns