AWS Summit ANZ 2022 - End-to-end MLOps for architects (ARCH3)

Learn how to design an end-to-end machine learning architecture, one step at a time, graduating from a simple model deployment to a complex multi-model strategy. This session aims to help architects working with data scientists and machine learning engineers to implement machine learning use cases. Prior knowledge of and experience with core AWS services is recommended.
Learn more about AWS webinar series in Australia and New Zealand at go.aws/3ChL0Y6.
Subscribe:
More AWS videos bit.ly/2O3zS75
More AWS events videos bit.ly/316g9t4
ABOUT AWS
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts.
AWS is the world’s most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers-including the fastest-growing startups, largest enterprises, and leading government agencies-are using AWS to lower costs, become more agile, and innovate faster.
#AWSSummit #AWS #AmazonWebServices #CloudComputing

[@jpzhang8290]

Best AI architecture talk so far. She is very hard-core and geared towards practical engineers, not idiots expecting to watch 2 hours of fancy PPTs.

[@gabesusman4592]

Small MLOps architecture completed @ 11:50 Medium MLOps architecture completed @ 15:45 Large MLOps architecture completed @ 20:40

[@mohddanish414]

I have been working in MLOps for last 3 years and it's the best resource available on youtube for architecting MLOPs process.

[@AWSEventsChannel]

🙌 😀

[@kishorkumar8080]

As a new MLOps engineer at NatWest, I found this video "AWS Summit ANZ 2022 - End-to-end MLOps for architects (ARCH3)" to be incredibly informative and helpful. The speakers provided a comprehensive overview of the key concepts and principles of MLOps, as well as real-world examples and case studies. I particularly appreciated their insights on how to implement MLOps in a large enterprise setting. Overall, I highly recommend this video to anyone who wants to learn more about MLOps or advance their career in this field.

[@buvaneswariramanan5010]

One of the best talks on MLOps! Kudos to the presenter for condensing so much information in such a short amount of time. The diagrams and the presenter's narration are super clear.👏👏👏

[@AWSEventsChannel]

Thanks so much for the feedback! So glad you like it! ☺️

[@mehditaslimifar2521]

Very interesting talk, is it possible to access the architecture diagrams?

[@gabrielnovak4507]

I was searching for some real MLOps architecture examples with commentary and there is almost no content like this on YT. Thank you for this video. Great density of information also :)

[@milanzinzuvadiya6447]

Great Talk on AI/ML architecture. kudos to the presenter. can anyone help me find those whitepaper links that were mentioned at the end of presentation?

[@MichaelBazik]

New to an ML ops team and can't agree more with the comments. Even in 2024, this is still a great primer for a devops guy getting asked to convert to MLOps. Thanks!

[@mmacasual-]

For Medium MLOps architecture what is the point in monitoring the model in the Staging environment? Monitoring production models are enough I guess. Please help me understand

[@beecee4756]

👍really intrigued my interest on MLOPS. Chose this over IPL live, that's how interesting this is for me :) Concise and to the point. Not a single minute wasted. :) :)

[@rajalamuri553]

thanks for this !! Can we implement most of this if we are deploying to an edge device?? Any major major changes.

[@visheshbhagat]

This is PURE GOLD!

[@kishorescm]

Your explanation of E2E ML architecture is fantastic! You managed to condense what would have taken me hours of reading and research into a clear and concise 23-minute presentation with great examples. Thank you for making such a complex topic so easy to understand. Perhaps sharing the Cloudformation scripts would save all massive amount of time to start exploring the services.

[@amazonwebservices]

Thanks for your feedback, Kishore! 😀 🙌 ☁️

[@andrewcbuensalida]

No link to the Cloudformation to deploy this?

[@Watchkcom-u9w]

Hall Charles Gonzalez Daniel Walker Richard

[@keslauche1779]

Great talk the best on MLOps on KZbin, can anyone give me resources to include offline or online feature pipelines with feature stores?

[@awssupport]

We're glad to hear you enjoyed this! ☁️ I've pulled together a few links that I believe you'll find useful: go.aws/4drS5VQ & go.aws/4doHhrm & go.aws/4doMi3h & go.aws/4drK9Ui & go.aws/4dNosOe. 🔗 I also encourage our re:Post platform for more helpful advice from our cloud experts: go.aws/aws-repost. 📮 ^TE

[@myselfandpesit]

Thanks for the video. You accurately talked about various points in the ML pipeline.

[@sahilchachere15]

Thank you for putting out this video, really helpful!!! Helped me clearly understand how MLOps can be leveraged at different hierarchies.

[@awssupport]

It's our pleasure, Sahil. Thanks for tuning in! 🤝 ^RB

[@ahsin.shabbir]

This is a great presentation and sounds lovely when looking at diagrams but most of the issues with AWS is that things rarely work as smooth in practice once you actually start using the services.

[@hanyelfouly3063]

Amazing video, But what about metadata store, and experimentation tracking I can't seem to find it on the architecture

[@ZyklonB-88]

oooo she looks like a fun person

[@harperjmusic]

This is by far the best MlOps yt vid out there

[@amazonwebservices]

That's great to hear! 😀

[@pklesta1]

Really good presentation and ideas for MLOps architectures. Thanks for sharing!

[@amazonwebservices]

Our pleasure! 🤩 🙌 ☁️

[@Ytsubray73]

Great explanation. Thanks for sharing

[@LuisMorales-bc7ro]

I loved this video

[@AWSEventsChannel]

So glad you loved it!😀 🙌

[@ndamulelosbg8887]

How does the approval process work? Do we need to approve each version that comes in or perhpas approve one model over some period e.g. a day?

[@awssupport]

Hello. Thank you for reaching out. We encourage you to reach out to our community of experts via re:Post here: go.aws/aws-repost. ^AM

[@captiandaasAI]

great!! thankyou!!

[@vindolanda6974]

kzbin.info/www/bejne/i5-kf2ZrnNpmiNk Why is it necessary to re-deploy the Production services every time a new model version is deployed?

[@awssupport]

Hi! I found this blog which may be helpful in answering this question: go.aws/3M36WLy. Alternatively, you can ask our re:Post community this question, so they can weigh in: go.aws/aws-repost. ^NR

[@reaganlopezmusic]

Thanks for the great presentation.

[@rakeshsukumar6226]

Can I get these slides?

[@RahmatKhan-sp7nb]

amazing!

[@amazonwebservices]

So glad you like it! 😀

[@anshulzade6355]

The amount and way of information that is presented in the video is amazing. Thanks!!

[@amazonwebservices]

Glad you liked it, Anshul! 🤝 😃 ☁️

[@ray811030]

I love the takeaway she gave in the end of talk

[@XShollaj]

This is amazing

[@kallingal1662]

Very well explained

[@AWSEventsChannel]

We're glad you found it helpful.😀

[@jamespan3417]

How can I download the slides of this talk?

[@awssupport]

Hello, for more information regarding this event, follow this link: go.aws/3ChL0Y6. ^RB

[@anthonyjoyes7248]

Superb!! Just what I needed🙂

[@amazonwebservices]

Great! 🙌 😀

[@kirityrapuru]

Awesome

[@siddharthshukla3557]

awesome.

[@TechnocracyInAction]

Where can I get slides?

[@awssupport]

Thanks for reaching out! I've sent off your request to our service team for review. 🧐 I don't have an ETA, but we'll update you as soon as we hear back. ^KS

[@awssupport]

Thanks for allowing me time to get back to you. I've received an update from our service teams, and we're unable to share the slides. You're welcome to screenshot the videos or sections you wish to save. Sorry about this one. ^KS

[@zortx233]

@@awssupport Can you at least post the URLs for the links shown at the end in the "Starting Your Journey" slide?

[@jamespan3417]

Awesome, indeed!