26 November 2018: Alex Milowski (Orange Silicon Valley)

Abstract

Enterprises face new challenges deliver features, at scale, from their data to continuously train models for AI and Machine Learning (ML) applications. Traditional "Big Data" systems, like Hadoop, have often been deployed to facilitate scaling parallel processing of large data sets but they have often failed to deliver in terms of cost, scale, and ease of use; especially for ML data pipelines. New frameworks, like Tensorflow, and cluster orchestrators, like Kubernetes, have put further pressures on "big data" infrastructure to evolve. Where do these new ML data pipelines get their data from? This talk will explore the future where "Big Data" has evolved to just "Data" existing in a multi-cloud fabric where metadata is the essential orchestrator.

Biography

Dr. Milowski is a researcher and entrepreneur and current is a Research Scientist at Orange Silicon Valley (a division of the French telecom Orange). He has been involved in the development of Web and its technologies since 1990, primarily focusing on data representation, algorithms, and processing data at scale. Recently, he has been working on improving the productivity of machine learning systems and the data scientists that use them. His focus has been how enterprises can use container orchestration technologies, such as Kubernetes, to offer greater efficiencies whilst preserving governance and privacy of the data within.

He received his PhD in Informatics in 2014 from the University of Edinburgh School of Informatics.

Nov 26 2018 -

26 November 2018: Alex Milowski (Orange Silicon Valley)

The Data Fabric: Evolving "Big" Data for AI Applications

IF 4.31/4.33