MLOps on Databricks: A How-To Guide

As companies roll out ML pervasively, operational concerns become the primary source of complexity. Machine Learning Operations (MLOps) has emerged as a practice to manage this complexity. At Databricks, we see firsthand how customers develop their MLOps approaches across a huge variety of teams and businesses. In this session, we will show how your organization can build robust MLOps practices incrementally. We will unpack general principles which can guide your organization’s decisions for MLOps, presenting the most common target architectures we observe across customers.
Combining our experiences designing and implementing MLOps solutions for Databricks customers, we will walk through our recommended approaches to deploying ML models and pipelines on Databricks. You will come away with a deeper understanding of how to scale deployment of ML models across your organization, as well as a practical, coded example illustrating how to implement an MLOps workflow on Databricks.
Connect with us:
Website: databricks.com
Facebook: / databricksinc
Twitter: / databricks
LinkedIn: / data. .
Instagram: / databricksinc

[@RedShipsofSpainAgain]

31:56 Part II begins (practical part of the talk)

[@lindadanlol]

Thank you for sharing this great demo!

[@HariKrishnaReddy7696]

Thanks a lot for your efforts. Its a great demo

[@prabhamelady1680]

Its a nice demo

[@adnanakbar2521]

Hi, is the code repository used in this demo publicly available?

[@georges7298]

Good session. One thing may need improving is that while showing diagrams, it's hard to read them - if the diagrams can be maximized to full screen, it would be much better

[@mwredfern]

does anyone have a link for this PPT? thanks

[@maazaouiyousri7806]

Thank you for the sharing. How can we get the ppt presentation ?

[@prpgift3313]

Thank you for making the great demo, so could I have a presentation file of this demo?

[@spillerme]

same question! pls reach me out if you found it
thx in advance

[@MrBemnet1]

So you wouldn't know how your model is performing on whole dataset until you deploy it to prod?

[@akhilsebastian3804]

Yes. Even I am stuck with the same doubt. The model training should be done on the latest data. It doesn't make sense to retrain model in dev, stage and prod.
Also the what is happening with the prediction code?

[@user-atad]

Seems very complicated tbh

[@brenok]

and what would you do? test in prod? :D

[@faiqkhan7545]

mlops engineer is a senior data role .

[@nitishbajaj1963]

Is there a code repository for this. Please let me know if I could learn from that. If yes, I would dm you my emailid Thanks