I like this video. As a data scientist, with data engineering responsibilities I see clear use cases for both tools. We use Sagemaker instead of DataBrew which should point to the differences. I would say that Glue studio is focused on data. There is no intention of doing anything with it except getting it, transforming it, and storing it (ETL). Jobs done. DataBrew is there for people who use data. For them, data is the tool, not the product. You want a dataset to get insights or train a model. The intention is different which also means that the skills and preferences of the users are different.

very informative sir... :)

By the way I love this video. I feel AWS has many redundant tools. And they should start to narrow their tools to limit confusion

Thanks for the video. It's so informative.

Thanks a lot. Very clear.

Uses of data lineage is that when we are collaborating with multiple teams and if we have multiple sources lineage would be a easier way to track where data is coming, transforming.

FYI: AWS Glue Studio also provides data preview btw. But I see that they have included it after this video was released.

Thank you for this informative video :) I would to submit a problem that i've experienced in Glue Databrew, if any of you can help it would be a blessing. Here's the situation : I would like to filter on a value of a column that isn't in the sample dataset. And I've found out that the recipe only focuses on the sample dataset. The fact that the sample is limited to only 5000 rows max, is preventing me from completing my recipe. Does somebody have an Idea on how to bypass the limits of the sample size ?

Ok thanks for the information. Now I understood, Databrew is more towards data preparation using ML. And data glue is more towards job processing using pyspark . And the similarity is both can do gui etl .

What is the difference between Databrew and sagemaker Data Wrangler?

Another key difference is that DataBrew doesn't offer custom transformation I rekon ?