Machine Learning in Energy: Forecasting, MLOps, and Business Impact

Welcome to a new episode of The Data Playbook Podcast by Dataminded. In this episode, Kris Peeters talks with Jean-Michel Begon, Senior Machine Learning Engineer at Luminus, about what it really takes to build, deploy, and maintain machine learning models in a business-critical environment. Together, they unpack how ML teams in the energy sector move from ideation to experimentation, industrialisation, and monitoring, and why production-ready machine learning is about much more than model accuracy alone. In this episode, you’ll learn: How Luminus uses machine learning for electricity consumption forecasting What a practical ML lifecycle looks like in a business team How to balance experimentation, standardisation, and production delivery Why GenAI and LLMs are useful tools, but not a replacement for engineering discipline Subscribe for more conversations on data strategy, data engineering, AI, cloud platforms, and modern data teams. Follow Dataminded on YouTube: ‪@Dataminded‬ Explore the full podcast series: The Data Playbook Playlist -    • The Data Playbook Podcast   Discover more podcasts, blogs, and webinars: Dataminded Resources - https://www.dataminded.com/resources Visit the Dataminded website: https://www.dataminded.com/ Chapters: 00:00 Introduction 00:38 Jean-Michel Begon & Luminus 01:18 Electricity Demand Forecasting with ML 03:12 Data Sources for Energy Models 04:09 Why Forecasting Matters for Energy Supply 06:12 Machine Learning Team at Luminus 07:40 The ML Lifecycle (Ideation → Production) 12:36 Experimentation and Model Evaluation 24:21 Monitoring Models in Production 43:19 LLMs and the Future of ML Work