SAP has provided a number of solutions to carry out AI/ML operations in a given SAP landscape. Each of these solutions have specific advantages and disadvantages. This blog touches upon SAP AI Core in SAP Business Technology Platform (BTP) and SAP HANA Cloud.
SAP AI Core in SAP BTP
SAP AI core is the most flexible and feature rich solution available in SAP BTP for implementing AI/ML scenarios. SAP AI Launchpad acts as an entry point for SAP AI Core to provide access for administration and monitoring of AI/ML cases in SAP BTP.
It also provides support for automation of end-to-end AI/ML scenarios by using MLOps.
MLOps enables easy maintenance and updates in AI/ML scenarios. It also ensures the usage of the latest dataset and AI/ML artifacts. The pipelines created using MLOps continuously synchronize the workflow logic in the Github and dataset present in the object storage. The AI/ML code and the workflow can also be modified very easily to pick up a docker image for a given AI/ML scenario.
One of the other advantages of SAP AI Core is that it allows the usage of a library of your choice for building AI/ML code. You can choose to make use of Python, R or any other library which suits your requirement. Being a part of SAP BTP Portfolio , SAP AI Core has an advantage due to the integration capabilities of SAP Integration Suite and features like Data Federation of SAP Datasphere.
Seamless integration of a hyperscaler Object Store with SAP AI Core allows the creation of a unified dataset from multiple SAP and non SAP sources. By utilizing hyperscaler Object Store for storing datasets and Docker for logic execution, SAP AI Core provides all the cloud benefits such as scalability, high availability, and cost-effectiveness.
SAP HANA Cloud
SAP HANA Cloud provides a single place to access, store and process all enterprise data in real time. It allows consumption of data from SAP HANA Databases by applications on SAP BTP or on-premise or any other cloud services using the standard SAP HANA clients.
This eliminates the need to load all the data into a single storage solution thus providing simplified data access. If all of the required data is available in SAP HANA Cloud, then it is advisable to carry out AI/ML logic execution in SAP HANA Cloud.
SAP AI Core is more suited for scenarios which require gathering data from multiple SAP and non-SAP sources. Also suitable for organizations which already have platforms to support generally used languages like Python, R or any other non SAP language. SAP AI Core makes use of external libraries, e.g. Scikit-learn in Python or mlr3 in R to carry out AI/ML operations.
However, the SAP HANA Cloud, SAP HANA database already has embedded libraries and functions such as PAL (Predictive Analysis Library) and APL (Automated Predictive Library). These libraries provide functions and procedures to carry out AI/ML operations in the SAP HANA Database. PAL and APL offer the advantage of executing AI/ML operations natively in the database without the need to retrieve the data into temporary tables.
Again, the triggering of AI/ML scenarios can happen outside the SAP HANA Cloud as well. For instance, these operations can be triggered from SAP AI Core through SAP’s Python library hana_ml.
If you have any questions about solutions available for AI/ML Scenarios in SAP BTP or SAP S/4 HANA, please connect with Mindset.
