If you weren’t able to attend our recent SkyFoundry and Energy Twin webinar, Small Data, Big Impact — How Main Meter Data Delivered 5% Energy Savings Across 60+ Buildings, the full session is now available to watch on demand. We hope you find the content valuable.
Machine learning is often discussed as a future promise in energy management—but how does it deliver value at scale, using real operational data? In this joint webinar, Energy Twin and SkyFoundry present a practical approach to applying Python-based machine learning with SkySpark® to analyze and optimize large building portfolios.
Using a real banking portfolio with 60+ branches as a case study, we demonstrate how individual ML models can be built for each building to enable:
Continuous anomaly detection and Monitoring-Based Commissioning (MBCx)
Identification of small but persistent performance deviations
ML-based benchmarking across portfolios
Estimation of building-level and portfolio-wide energy saving potential
Using only 15-minute main meter data, this approach delivered over 5% portfolio-wide energy savings during the first year of operation, with clear site prioritization and transparent verification.
The webinar demonstrates how SkySpark provides a scalable and flexible foundation for advanced ML workflows, and how combining it with domain-specific energy expertise turns raw data into actionable, verifiable energy insights.
Debbie Bretches Thu 18 Dec 2025
If you weren’t able to attend our recent SkyFoundry and Energy Twin webinar, Small Data, Big Impact — How Main Meter Data Delivered 5% Energy Savings Across 60+ Buildings, the full session is now available to watch on demand. We hope you find the content valuable.
Watch the recording here: https://bit.ly/49CeX5q
Machine learning is often discussed as a future promise in energy management—but how does it deliver value at scale, using real operational data? In this joint webinar, Energy Twin and SkyFoundry present a practical approach to applying Python-based machine learning with SkySpark® to analyze and optimize large building portfolios.
Using a real banking portfolio with 60+ branches as a case study, we demonstrate how individual ML models can be built for each building to enable:
Using only 15-minute main meter data, this approach delivered over 5% portfolio-wide energy savings during the first year of operation, with clear site prioritization and transparent verification.
The webinar demonstrates how SkySpark provides a scalable and flexible foundation for advanced ML workflows, and how combining it with domain-specific energy expertise turns raw data into actionable, verifiable energy insights.