Freelancer profile translated to English.

Description

With 3 years of experience in data engineering and analysis, this profile has a perfect command of the Azure ecosystem and Big Data technologies. An expert in PySpark and Python, they have developed complete data processing solutions on Azure Databricks and implemented Delta Lake architectures. Their background in mathematical modeling and Machine Learning skills allow them to design high-performance predictive models. Versatile, they combine technical expertise with business vision to transform data into strategic insights.

Languages

French
Native or bilingual
Arabic
Native or bilingual
English
Fluent
German
Basic

Workplace preferences

Can work on-site

Clermont-Ferrand (up to 50km)

Michelin
Data Engineer
RAW MATERIALS INDUSTRY
October 2022 - April 2025 (2 years and 6 months)
Clermont-Ferrand, France
Context:

Integrated into an international Data team (Agile Scrum) on a project to modernize and industrialize data flows for Michelin's industrial and business departments.

Collaboration with teams spread across several countries, with technical meetings and exchanges in English.

Missions:
Design, development, and production deployment of ingestion and transformation pipelines in an Azure DataLake based on Delta Lake.
Data integration via Azure Data Factory (SQL databases, files, APIs) and real-time stream ingestion with Kafka for streaming data.
Advanced development on Databricks (PySpark, Python, SQL): creation, optimization, and maintenance of critical jobs (batch & streaming).
Configuration and optimization of Databricks clusters, improvement of PySpark code to reduce execution times and costs.
Implementation of production monitoring: tracking Databricks jobs via Power BI and automated alerts with Power Automate to detect failures in real-time.
CI/CD automation via Azure DevOps, with code versioning using Git, execution of unit tests, multi-environment deployments, and rollback strategies to secure production.
Technical support and incident resolution: analysis of Databricks logs, quick fixes, and assistance to business users to ensure flow availability.
Contribution to the governance and documentation of pipelines, with a focus on quality, performance, and reliability.
Reduction of data processing time by 30% through optimization of PySpark jobs and the use of Delta Lake partitioning features
Microsoft Azure Databricks Data Engineer Big Data PySpark
Groupe Centre France
Data Analyst & BI
PRESS AND MEDIA
September 2021 - September 2022 (1 year)
Clermont-Ferrand, France
Context:

Integrated into the Business Intelligence (BI) department of a regional press group, in direct interaction with the marketing, distribution, and finance departments.

Strategic project to migrate the decision-making environment from JasperSoft to Tableau Software, in order to modernize steering tools and improve user experience.

Missions:
Conduct an audit of JasperSoft usage: analysis of existing reports, understanding user interactions, identification of critical KPIs for business departments.
Present results to decision-makers to define a prioritization of developments under Tableau, aligned with strategic needs (subscriptions, sales, advertising revenue).
Design and implement new interactive dashboards in Tableau, integrating data from SQL Server, PostgreSQL, and MySQL via Denodo for data preparation and virtualization.
Perform numerous SQL queries on Microsoft SQL Server to extract, clean, and consolidate data used in reports and dashboards.
Develop and maintain existing Jasper reports in Java to support the transition phase.
Conduct ad-hoc analyses and write functional specifications, conduct user acceptance tests (UAT), and train end-users
SQL Data visualisation Tableau software Denodo
Coffreo
Data Scientist
HUMAN RESOURCES
April 2021 - September 2021 (5 months)
Aubière, France
• Development of a predictive customer consumption model to anticipate usage volumes and optimize resources.
• Enrichment of internal data with open data sources (regional socio-economic data, weather data).
• Design of a complete pipeline: preprocessing, feature engineering, training, and evaluation of Machine Learning and Deep Learning models (neural networks).
• Documentation of methodologies and knowledge transfer to internal teams.
• Improvement of forecast accuracy by 25% compared to previously used statistical models.
• Provision of an interactive dashboard in Python/Plotly for forecast visualization.
Machine learning Python Data science Forecasting Deep Learning