SidAli CHORFI

Worked on the Leclerc transformation program across several SCA entities: Scapartois, Scacentre, Scanormande, Scapest, Scaso, Scapalsace, on various data-related initiatives: • Designed and implemented a data platform on GCP dedicated to data analytics, covering the full data lifecycle from ingestion to exposure in BigQuery. • Deployed and configured required GCP services to support data processing pipelines. • Technologies: Terraform, GCP (BigQuery, Cloud Composer / Airflow, IAM, Cloud Storage, Datastream, Artifact Registry, Dataplex), dbt, Python. • Led and facilitated workshops on data platform implementation, covering key GCP services and CI/CD best practices. • Implemented a batch replication pipeline from a Progress OpenEdge database to PostgreSQL, supporting initial (full) and incremental (delta) loads including insert, update, and delete operations. • Technologies: Python, Alembic, PySpark, crontab, SQL, SSH.

Worked on two projects at Veolia: 1. Developed various KPIs based on real-time water sensor data ingested via a Kafka cluster. • Data was stored in AWS S3 data lake buckets, transformed using PySpark, and finally exposed to Data Analysts for analysis and reporting. 2. Devloppment of a Proof of Concept about calculation of water trace, using Neo4j and Docker.

Build phase: • Ingested and pre-processed data from multiple sources and stored it in AWS S3 buckets following a Diamond architecture (Bronze, Silver, Gold). • Monitored data transfer pipelines using Grafana. • Transformed Bronze data into Gold datasets using PySpark and Spark SQL. • Exposed data through AWS EMR jobs and Airflow DAGs. • Deployed data transformations to production at the end of each sprint using GitLab CI. Run phase: • Monitored daily data loads to ensure data quality and pipeline reliability. • Handled production incidents and alerts, including root cause analysis and debugging. • Ensured clear communication around incident management via email and ServiceNow. • Managed access control to S3 buckets for data consumers using IAM and Terraform.

Working in an 𝗔𝗴𝗶𝗹𝗲 Environment: - 𝗘𝗘𝗚 𝘀𝗶𝗴𝗻𝗮𝗹𝘀 𝗖𝗹𝗮𝘀𝘀𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 using Deep Learning techniques Preprocessing techniques over Signals. - 𝗧𝗲𝘅𝘁 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻 applied to Political speeches using 𝗡𝗟𝗣 techniques and DeepLearning 𝗟𝗦𝗧𝗠. - 𝗗𝗮𝗺𝗮𝗴𝗲𝗱 𝗰𝗮𝗿 𝗜𝗺𝗮𝗴𝗲 𝗦𝗲𝗴𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗰𝗹𝗮𝘀𝘀𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 using 𝗖𝗡𝗡: RetinaNet, Mask-RCNN. 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀: Python, NLP, DeepLearning, LSTM, CNN, Scikit-learn, tensorflow, Keras, Pytorch, Preprocessing on Data, Colab, Jupyter.