End-to-end MLOps with Vertex AI

Understanding ML pipelines → goo.gle/3FfsjlR
Codelab → goo.gle/3zNFFTL
Want to build a system that continuously evaluates and updates your model to reflect changes in its environment? In this episode of AI Simplified, we’ll walk you through a demo of end-to-end MLOps workflow using Vertex AI pipelines. Watch to learn how MLOps can help you build a system that reflects changes in the environment!
Chapters:
0:00 - Intro to MLOps workflow
2:30 - MLOps Workflow demo
6:53 - Compile and run end-to-end ML pipeline
8:00 - Summary
Watch more episodes of Getting Started with Vertex AI → goo.gle/AISimplified
Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech
​
#AISimplified
product: Cloud - AI and Machine Learning - AI building blocks; fullname: Priyanka Vergadia; re_ty: Publish;

[@sukanya4498]

One of the Best videos I found on introducing MLOps to a Beginner within 10 minutes ! Thank you !

[@googlecloudtech]

You're very welcome!

[@peterpaul2773]

Awesome! Thank You

[@TomaszWawa]

Great, thanks.

[@vineetatewari8808]

Thanks.

[@itgdata]

Great!

[@user-wr4yl7tx3w]

What if you need to retrain the model after deployment? Can Vertex AI monitor that situation? Is there a video on that aspect?

[@RajivMukul]

Thanks for the tutorial.Can you please tell me in which scenarios python scripting is used in MLOps workflow apart from building a model .Also in which cases we use bash.Basically want
to know how much scripting is needed in MLOps apart from building a model.

[@mrc7701]

Priyanka ... thank you for what you do. It is simply remarkable on how you break down these concepts into easy-to-digest chunks.
A question for the term "end-to-end" AI pipeline ..... can this end-to-end involve steps spanning multiple clouds ? for instance, an end-2-end workflow that involves training model on GCP and Deploying model on AWS (or any other cloud for that matter) and GCP being the sole controller of this end-2-end workflow. Would love to hear your thoughts .. thanks again.

[@infiniteloop5449]

push your model to an SNS topic endpoint in AWS. alternatively have your AWS service long-poll on a google storage bucket.

[@VigneshVicky-cn8ek]

Cool one

[@jacobmiller38]

Is there cli tools for vertex?

[@greendsnow]

i regret being 1 year too late to these video series.

[@asadfaizi]

Please share the link to previous video where you introduced MLOps

[@laxsfogrusna1979]

Priyanka's Intro to MLOps - czcams.com/video/Jrh-QLrVCvM/video.html

[@kalyanbdlive2837]

Why Google update complite?

[@AmitYadav-ov9ie]

Where is the Monitoring part? Where is the Pipeline scheduling part? Where is the monitoring or logging the metrics or Drift part?

[@vigneshv9584]

Mam i am BSC chemistry degree complete but google company work tell me

[@SaiKalyan101]

I would have been if you demonstrated without autoML.

[@kadapa-rl6jg]

Handling vertex AI with respect to worst roles and permission is hard to handle