Guillaume Geoffroy

Designed and managed data pipelines for energy consumption and billing data.

Developed a Python library to structure ETL workflows, including complex PySpark transformations on time-series data. Worked on VSCode with Github Copilot.

Industrialized and managed Databricks jobs, with scheduling through Apache Airflow and data storage on S3 using Delta Lake format.

Orchestrated CI/CD with Azure DevOps and IaC deployments with Terraform across Databricks environments (dev, preprod, prod).

Integrated GitHub Copilot into the development workflow for code generation, refactoring, and pull request review support.

Built a Data Quality framework within the library, implementing checks for duplicates, overlaps, and completeness. Used Docker Image for unit testing / functional testing.

Performed data analysis and developed dashboards with Databricks.

Analyzed weather sensor data to anticipate late frost events.

Led the project autonomously, coordinating with multiple stakeholders.

Analyzed time-series data from weather sensors and developed solutions on Linux using Python (Pandas, Matplotlib, TensorFlow) and MySQL.

Designed a Proof of Concept and built a Deep Learning model (CNN/LSTM) to

estimate dew point at parcel level.

Predicted aircraft failures for Airbus and airline operators.

Filtered, analyzed, and visualized multi-source aircraft sensor data, including model development and alert monitoring.

Developed a Python library dedicated to model development, built on complex

PySpark transformations.

Industrialized Big Data models using internal DevOps tools within a continuous

integration framework.

Used the internal CodeWorkbook ETL for model prototyping and validation.