Forecasting based proactive optimization of cloud resources:

The workshop uses practical examples to demonstrate the innovative approach to proactively optimize cloud resources based on dynamic and anticipated component usage. Application workload prediction is provided to a state-of-the-art machine learning based solver. This calculates an optimal deployment plan for the application to forecast future demand. State-of-the-art methods such as ES-Hybrid and innovative Monte Carlo Tree Search based solvers are used to predict an optimal solution.

Paweł Skrzypek is an experienced architect of IT solutions, especially in the field of big data processing and machine learning solutions. In the years 2006 – 2015, he co-created the architecture of IT systems solutions for the biggest companies in Poland. In the years 2016-2019 he carried out projects in the area of Cloud Computing and AI and deployed one of the most advanced AI solutions for the investments industry. He is currently the Technical Director of the MELODIC multicloud optimization and management platform.

Marta Różańska is a PhD student at the University of Oslo and a researcher at 7bulls. She has been working on cloud computing optimization for both commercial and research projects for 4 years, it is also a topic of her PhD thesis. Marta is leading the development of the proactive adaptation function in the ongoing MORPHEMIC project.

Alicja Reniewicz is a technical leader of MELODIC development team and a full-stack developer (in technologies related to web applications: Java, Spring, Angular), code integration and maintenance, use of cloud services also in programming. Alicja presents the MELODIC platform at various conferences to potential customers from a technical point of view.

This post is also available in: German