2025-09-26 –, Orbit
This talk explores how data science helps balance energy systems in the face of demand volatility, generation volatility, and the push for sustainability. We’ll dive into two technical case studies: churn prediction using survival models, and the design of a high-availability real-time trading system on Databricks. These examples illustrate how data can support operational resilience and sustainability efforts in the energy sector.
Energy may seem simple: pay a provider, receive electricity, not much data involved. But the reality is far more complex. With multiple producers, consumers, and grid operators, and the push for sustainability, balancing energy becomes a challenge in which data plays a big role.
This talk explores the complexities of the energy sector, focusing on the imbalance problem: energy in must equal energy out on the grid at every single moment. We’ll dive into two areas where data helps Eneco keep the grid stable while supporting society’s shift toward more sustainable energy use:
- Dealing with demand volatility: Using churn prediction to improve long-term planning and customer retention
- Dealing with generation volatility: Building a mission-critical, high-availability system for real-time energy trading data
Pablo Estevez leads Data and Machine Learning in Eneco’s Energy Trading teams. For more than ten years he has worked across tech, taking on both hands-on and leadership roles in Machine Learning and Data Science at companies like Booking.com and Meta
Manolis Manousogiannis is a Senior Data Engineer in Eneco's Energy Trading team, specializing in distributed data processing with Spark and Databricks. He holds a background in Computer Science and brings extensive experience in building scalable data solutions.