Time Series Forecasting at Scale – Lessons learnt from more than 10000 training jobs

Time series forecasting production workloads are challenging by nature. More than often individual simple statistical models like ARIMA or Prophet tend to give sufficient performance to not consider more sophisticated approaches. However, maintaining and retraining so many individual models can become a challenge. In this session you will learn what Siemens Energy is doing to solve this challenge and its future plans to optimize their current solutions

• Forecasting strategies (recursive, direct, mimo and hybrid)
• Metadata oriented machine learning
• Complex cloud machine learning pipelines
• Serverless active model monitoring

————————————————————————————————————————————————————

Edgar Bahilo Rodríguez – Lead Data Scientist | Siemens Energy

Edgar Bahilo Rodriguez is the lead data scientist in the industrial applications division of Siemens Energy, his main experience is in applying ML for time series analysis and forecasting but extends to other knowledge areas like anomaly detection and NLP. Edgar is the product owner of SEER, the company’s most notable end to end data scientist platform. He is also the architect of the data science life cycle in industrial applications, and he has strong experience implementing and architecting ML workloads in the cloud (AWS).  He co-leads the cross topic of ML Methods and Engineering at Siemens Energy.

Mohamed Ahmed – Senior Data Scientist | Siemens Energy

Mohamed Ahmed is a Senior Data Scientist working for Siemens Energy in the industrial applications division. He has worked in many time series forecasting problems and computer vision use cases. Recently he has been involved in the first IoT use cases using machine learning in industrial applications. Experienced with developing end to end ML solutions on cloud infrastructure (AWS), using MLOps best practices. He is also part of the ML Methods and Engineering cross topic at Siemens Energy.