Webinar: Hopsworks, Data-Intensive Machine Learning with a Feature Store

Data Science Serbia
Data Science Serbia
Public group

Online event

This event has passed

Details

Data Science Srbija Webinar
Hopsworks, Data-Intensive Machine Learning with a Feature Store

Hopsworks is an open-source platform for developing and operating machine learning models at scale that is now available as a managed platform on AWS at www.hopsworks.ai. Hopsworks includes the industry’s first Feature Store for machine learning that is usable as a stand-alone system from Databricks, Sagemaker and other ML platforms. This talk will introduce some of the new ML infrastructure trends that are enabling faster and lower-cost deployment of models: the Feature Store to manage and share features across organizations, feature engineering and training pipelines - including using notebooks in ML pipelines, provenance and reproducibility for ML pipelines, and automated monitoring of models in production.

Speaker bio:
Theofilos Kakantousis is a Data Engineer and co-founder of Logical Clocks (https://www.logicalclocks.com/). He holds an MSc in Distributed Systems from KTH Royal Institute of Technology and he frequently gives talks on Hopsworks. He has presented Hopsworks at venues such as Strata San Jose/New York, BigThings Madrid, BigData Moscow and BeamSummit Berlin.

All applicants will be able to attend the webinar via the zoom application. The access link: https://us02web.zoom.us/j/89304807483