
Building Well-Architected Machine Learning Solutions on AWS with Phil Basford
ML Platform Podcast
00:00
AWS SageMaker Batch API
Phil: I would abandon the app sees real-time nature lists. And then I would split it up, split back up, batch up into a couple faces. One that gets all the data as a dump out of the database the most efficient way and maybe storing it in something like S3. Then I would then do a batch transform with that model and then with the results in S3 load it back into the database as a kind of a subsequent step. The other option, and you'd have to do into the exact detail, is if you could bring that into something like Redshift,. So that's a way of then being able to run that directly on that data sets. That
Play episode from 34:27
Transcript


