Mohamed Boukechouch

• Development of an automated Python batch pipeline orchestrating the loading of over 10 million records per day into a BigQuery Data Warehouse, ensuring scalable and performant ingestion on the cloud.

• Design of an optimized star schema reducing query latency by 60% and facilitating multi-dimensional analysis via dbt.

• Creation of interactive dashboards connected to BigQuery, integrating business KPIs and offering a clear and actionable vision to decision-makers.

• Implementation of Airflow orchestration automating daily data flows.

Tech stack : GCP(BigQuery), Python, SQL, dbt, Airflow, Jira

Automation of Machine Learning (MLOps) Pipeline PFE (Bachelor's Degree)

Objective: Orchestrate the ML model lifecycle to reduce manual errors.

— Design of an automated workflow with n8n connecting data ingestion and processing.

— Development of Python scripts for cleaning (Pandas) and training models (XGBoost, Random Forest).

— Implementation of evaluation metrics (Accuracy, F1-Score) to validate prediction performance.

— Result: Functional pipeline capable of triggering automatic retraining on new data.