Comprehensive and Detailed Explanation From Exact Extract: Building custom AI solutions directly on hyperscalers introduces complexities such as implementing security measures to ensure compliance and data protection, integrating identity management for secure access control, and managing GPU clusters for scalable AI training and inference. These challenges arise from the need to handle infrastructure, integration, and operations manually, which SAP BTP mitigates by providing a standardized, hyperscaler-agnostic platform.
Exact extracts supporting this:
"Transitioning to a hyperscaler can help, but may still require dealing with integration and security complexities."learning.sap.com
SAP AI Core is "designed to manage the execution and operations of AI assets in a standardized, scalable, and hyperscaler-agnostic manner," implying complexities like GPU management on hyperscalers.help.sap.com community.sap.com
Integration challenges include "typical integration challenges and the integration journey in a multi-cloud environment," encompassing identity management.community.sap.com
Other options are incorrect because:
Option C: While selecting an appropriate LLM is important, the complexity is not specifically "choice of the wrong LLM" but rather model management; SAP emphasizes broader operational issues.
Option E: Data replication is a data management task but not highlighted as a primary complexity in hyperscaler AI builds; focus is on security, integration, and infrastructure.
References from Positioning SAP Business AI Solutions as part of SAP Business Suite documents or Study Guide: From SAP Learning Journey "Boosting Your Cloud Transformation Journey with SAP Business AI and Generative AI," units on building custom AI solutions and positioning SAP Business AI in cloud transformation. Supported by SAP Help Portal for SAP AI Core and community blogs on generative AI with SAP, aligning with C_BCBAI_2502 materials for comparing hyperscaler vs. SAP BTP complexities.
Submit