Wanaim Essaadia

– Designed and developed an intelligent monitoring system for anomaly detection on a limestone industrial conveyor belt. – Implemented Variational Autoencoders (VAE) trained on normal images to automatically detect anomalies using reconstruction-based methods. – Processed real-time video streams from a VMS (.dav format) using FFmpeg, performing frame-by-frame analysis and anomaly detection based on LPIPS distance. – Automated the storage of anomalous images, integrated alert generation, and developed a professional visualization interface. – Deployed the solution in an industrial environment with strict performance, reliability, and lightweight model constraints.

– Designed and implemented end-to-end ETL pipelines integrating data from multiple sources (relational databases, flat files, APIs), ensuring high data quality through advanced cleansing and transformation. – Developed and optimized interactive Tableau dashboards to monitor business activity, applying data visualization best practices (UX/UI, clarity, storytelling) to enhance user adoption. – Improved dashboard performance and reporting reliability through optimized calculations, aggregations, and automated refresh processes. – Collaborated with business stakeholders to translate requirements into scalable BI solutions, leveraging SQL and Python for data preprocessing and automation.