About Michael van Duinen

Architecting and operating production models and data pipelines for North American energy markets.

• Gaslytics. Lead supply and storage development, contribute to MLflow-managed demand models, and connect the system’s model workflows. Designed storage modeling across approximately 30 U.S. and Canadian geographies.
• Cloud modernization. Architected and deployed the NGL and coal modeling pipelines on Databricks, translating legacy R and MATLAB systems into production Python and PySpark workflows.
• Integrated Supply Forecast. Built long-term production forecasts across approximately 150 U.S. and Canadian reporting areas, from well-data pipelines and decline curves to analyst scenario tools. Rebuilt the MATLAB system in Python, pandas, and Streamlit.
• Production ownership. Independently maintained the legacy Bentek MATLAB system for three years alongside Gaslytics development, covering performance analysis, retraining, deployment, and enhancements for pipeline flows and LNG facilities.

Completed a three-year rotation across North American energy analytics, focusing on natural-gas infrastructure and upstream production.

• Infrastructure analysis. Owned a weekly report on pipeline and LNG expansion projects across Texas and the Southeast, alongside daily and monthly market analysis.
• Production research. Produced the monthly Northeast U.S. oil and gas report, analyzing drilling and production activity in the Marcellus and Utica basins.