ISMAIL ISMAILI ALAOUI

Developing, and maintaining data pipelines to efficiently process billions of vehicle trips data records, addressing business needs for analysis and vehicle performance tracking. Key Achievements: - Implemented Spark in Python to distribute data processing for extensive vehicle trips datasets. - Developed Spark processes in Python for transforming/aggregating vehicle data, feeding Power BI dashboards for vehicles performance monitoring. - Automated ETL processes on billions of connected vehicle trips records, reducing monthly manual workload. - Engineered APIs/libraries with Python/PySpark that support the daily team analysis, simplifying data post-processing. - Created pipelines in Azure Data Factory to migrate on-premise data to the cloud.

• Contributed to the development of ANPR (Automatic Number Plate Recognition) models tailored for Moroccan license plates. • Preprocessed and annotated thousands of vehicle images to train and evaluate deep learning models. • Evaluated multiple CNN architectures (YOLOv5, EfficientDet) to optimize plate detection accuracy under diverselighting and angle conditions. • Conducted error analysis and proposed enhancements to improve OCR precision on Arabic and French characters.