Mohamed El Mokhtar El Waghf

Missions and Responsibilities – Data Team at Side (A digital temporary work agency, part of the global Randstad Group) - Design and maintain robust data pipelines, ensuring reliable batch (Airflow DAGs) and streaming ingestion (via Kafka) into the Data Warehouse (PostgreSQL) for - analytical use. - Integrate and manage new external data sources (APIs, BigQuery, etc.). - Migrate legacy APIs from Flask to FastAPI. - Collaborate closely with the Business Intelligence (BI) team to deliver performant and business-oriented data solutions. - Develop and implement optimization strategies to improve Looker performance and scalability. - Actively support the full transition to dbt for data transformation and modeling. - Enhance CI/CD pipelines to improve automation and deployment efficiency. - Optimize queries, processes, and APIs that impact business and product performance. - Contribute to team knowledge sharing through dedicated workshops and documentation. - Contribute to the migration toward a new billing tool. Technologies: Python, PostgreSQL, Looker, Airflow, Kafka, dbt, Git (CI/CD), Docker, AWS S3, AWS RDS.

Au sein d’Orange, j’ai intégré deux équipes en tant que Data Engineer : Proxydata : - Collecte de données en mode batch ou streaming - Transformation des données si nécessaire - Mise à disposition des données aux partenaires internes à travers des dashboards ou des gisements de données. Profiling : - Analyser des centaines de millions de lignes de logs de consommation de la télé Orange pour calculer des profils de consommation. - Fournir les données nécessaires pour nos clients afin de créer des markers utilisés dans la recommandation de contenu ou d’autres besoins par nos clients internes et externes. - Assurer la qualité des données livrées ainsi que le respect des contraintes GDPR. En suivant la méthodologie Scrum, j’ai aidé les deux équipes à migrer les dataflows et workflows existants de l’on prem vers le cloud GCP. J’ai également participé au développement des nouveaux dataflows, des nouvelles fonctionnalités dans les workflows existants, ainsi qu’à la correction des bugs et à l’amélioration de la qualité du code et des données fournies. Compétences : Gitlab (Pipelines CI/CD) · Traitement et analyse de données · Scrum On-Premise : Nifi, Spark, HDFS, S3, Kafka, MapReduce, Scala, Bash, Pig, Hive, Oozie, Hadoop, Java, Grafana GCP : BigQuery, GCS, Airflow (Composer), PubSub, Python, SQL, DBT, DataStudio, Looker, Java, Dataproc

Within Orange, I joined two teams as a Data Engineer: Proxydata : - Data collection in batch or streaming mode - Data transformation if necessary - Provision of data to internal partners through dashboards or data pools. Profiling : - Analyze hundreds of millions of lines of Orange TV consumption logs to calculate consumption profiles. - Provide the necessary data for our customers to create markers used in content recommendation or other needs by our internal and external customers. - Ensure the quality of the delivered data as well as the respect of GDPR constraints. By following the Scrum methodology, I helped the two teams to migrate the existing dataflows and workflows from on prem to the GCP cloud. I also helped to develop new dataflows, new functionalities in the existing workflows, as well as to fix bugs and improve the quality of the code and the delivered data. Skills : Gitlab (Pipelines CI/CD) · Traitement et analyse de données · Scrum On-Premise : Nifi, Spark, HDFS, S3, Kafka, MapReduce, Scala, Bash, Pig, Hive, Oozie, Hadoop, Java, Grafana GCP : BigQuery, GCS, Airflow (Composer), PubSub, Python, SQL, DBT, DataStudio, LookerStudio, Java, Dataproc

Aldes aeraulique has connected products where data are sent and stored in a cloud server. The goal of this internship is to analyze this time-series data and use machine learning methods on it. The principal tasks are : - Optimize retrieving data : retrieve all data from the Microsoft Azure cloud for all products in a reasonable time - Present some global statistics and analysis - Analyze products by typology and detect some tasks that can be treated with machine learning methods - Build an anomaly detector model for TFLOW typology - Solve other tasks defined in analysis step

The goal of this project was to build an emotion, age and gender detector. The first part of the project was to write a bibliographic research on object detection models. Three models were used in the realization of the final product: a face detection model and then three other classification models (emotion, age and gender) which take as input the output of the first model. The developed detector takes as input either an image, a video or a stream (webcam). A version (not the final one) was deployed online at https://eag-detector.herokuapp.com/.