Ludovic Bénard

Context: Enhancing data from waste sorting centers. • Data IOT processing to delta lake layers according to medaillon architecture (bronze to gold) • Designed and developed automated ETL pipelines with Azure Synapse Analytics and PySpark • Developed and implemented data unit tests to improve the data quality (automated tests through ETL and JSON Schema library) • Azure DevOps: CI / CD flow for Azure Synapse

Context: Industrialization of the anonymization tool for the PRISME cluster. Roles: • Anonymization of sensitive project data using Apache Nifi ETL and Python scripts • Imports / Exports of database dumps via SFTP • Projects scoping (functional and technical) and advices on RGPD rules • Gitlab CI /CD: Designed flow from dev to prod environment. Industrialization of anonymization solution with Docker images and Kubernetes

Context: Implementation of a handbag model recommendation system using computer vision. Roles: • Developed Snowflake stored procedure to get handbag images into AWS S3 bucket. • Developed PowerBI report to analyze handbag caracteristics • Image preprocessing / Feature engineering • Trained and deployed 4 Deep Learning models in EC2 containers via Amazon SageMaker and Jupyter Lab

Context: Subcontractor recommendation. Roles: • Developed PySpark scripts to process data from bronze to gold layer with Azure Databricks • Designed and developed automated ETL pipelines with Azure Data Factory • Developed a clustering algorithm with Natural Language Processing to classify construction sites • Deployed a model to an Azure Kubernetes Service • Developed PowerBI reports for data analysis and classification quality

Context: Account migration follow-up Roles: • Analyzed of existing data in the production database • Developed with DAX queries and Automatized daily emailing and refresh of the PowerBI report

Artificial intelligence bootcamp