Accelerating Machine Learning App Development with Kubeflow Pipelines (Cloud Next '19)

Building production-grade machine learning applications that run reliably and in a repeatable manner can be very challenging. Machine learning systems often need to orchestrate many steps, from data pre-processing and feature engineering, to model training, evaluation, and deployment. Teams need a structured way to develop, orchestrate, and run such multi-step pipelines, without sacrificing rapid prototyping and experimentation.
Find out how running Kubeflow on Google Cloud helped GOJEK to dramatically accelerate the speed at which they could deliver machine learning applications into production.
Accelerating Machine Learning App Development → bit.ly/2TZfO60
Watch more:
Next '19 ML & AI Sessions here → bit.ly/Next19MLandAI
Next ‘19 All Sessions playlist → bit.ly/Next19AllSessions
Subscribe to the GCP Channel → bit.ly/GCloudPlatform
Speaker(s): Anand Iyer, Willem Pienaar
Session ID: MLAI211
product: Cloud - General; fullname: Anand Iyer, Willem Pienaar; event: Google Cloud Next 2019;

[@adityagautam4974]

AI hub is mind-blowing. It is like each some code by people around the world is shared and re-usable. This will make life really easier for all ML engineers and scientist.

[@kityeung6229]

It is a really good introduction presentation about Kubeflow. Thank Anand and Wellem.

[@hitesh1907]

Hi Willem/Anand, Thanks for the session. Can you please share the notebook example(Allocation) that was shown by Willem. It will be helpful to do these steps on own.

[@darkwingduck42]

Is it pronounced Cube flow or KOOB flow?

[@zzz321yyy3]

I thought it was Cube flow until i saw this video