Building Model in cloud

[Narasimha]

How does building a model in SAP is different than building a model on GCP / AWS or Azure ? If SAP is hosted on cloud on either GCP, AWS or AZURE what is my advantage of ETL data to a a Data Lake on clouds vs SAP and do ML

[MarioDeFelipe]

Hi Narashima

I believe the first question is for HANA Cloud. There are fundamental differences between Amazon, Microsoft, Google, and SAP regarding DBaaS and analytical tools. Without the intention to compare all of them, I would say (if Denys Van Kempen does not immediately correct me) that SAP HANA Cloud would be a great fit if your company is oriented towards SAP data sources, SAP BTP, and its analytical services like SAC.

Amazon or Google would be a great fit if heterogeneous data sources come into play; I would say Google tops with BigQuery and Looker. Amazon has been for many years the platform of choice with the combination of S3 and Redshift.

If you want to know about HANA Cloud Development, like creating tables, views, calculation views, accessing data from data lakes via virtual tables, and more, I encourage you to look at SAP Cloud Application Programming Model.

https://help.sap.com/docs/BTP/65de2977205c403bbc107264b8eccf4b/00823f91779d4d42aa29a498e0535cdf.html... your second question, if SAP is hosted on cloud on either GCP, AWS or AZURE what is my advantage of ETL data to a a Data Lake on clouds vs SAP and do ML.
With SAP you will always have fewer requirements to ETL, but still, there might be a need to do it; just mentioning available methods (and not advantages) I give you a screenshot of the different ways to extract data on AWS, just to name an example.

Source, AWS

Multiplied by 4, not to mention that SAP can be hosted on Microsoft and the required data lake on Amazon S3.

Conclusion for the second question, I would rather on the business requirement, then let's choose what technology is better for each particular case