The joint “Databricks Lakehouse for SAP” solution—commonly called SAP Databricks in SAP partner material—lets customers stream or replicate operational data from SAP systems, blend it with third-party sources inside the Databricks Lakehouse, and directly apply Databricks’ advanced AI/ML runtimes (MLflow, AutoML, etc.). SAP’s own press release on the partnership states that the offering “enables organizations to combine SAP and non-SAP data to build advanced AI and machine-learning models.” None of the other listed SAP products supply an integrated lakehouse runtime for large-scale model training; they either manage models, visualize data, or provide data warehousing only.

A. SAP AI Launchpad – Governs lifecycle and deployment of existing models; it does not supply data-engineering or large-scale training across disparate sources.

B. SAP Analytics Cloud – Purpose-built for BI planning and visualization; contains only lightweight smart-features, not an AI/ML development runtime on combined data.

C. SAP Datasphere – Supplies a unified semantic layer and integration, but modelling must be done in external engines such as Databricks or SAP HANA Cloud ML.

1. SAP SE News Center

“SAP and Databricks Partner to Help Customers Simplify Data Landscapes and Unlock the Power of Business Data

” 08-Mar-2023

para. 3 & 4 (states joint lakehouse “enables advanced AI and machine-learning models on SAP and non-SAP data”).

2. Databricks & SAP

“Databricks Lakehouse for SAP — Solution Brief

” v1.3

2024

pp. 2-3 (architecture diagram showing SAP + third-party data feeding ML pipelines).

3. SAP Help Portal

“SAP AI Launchpad – Product Overview

” Section ‘Capabilities

’ 2024-05 (describes model operations

no data-prep/training).

4. SAP Help Portal

“SAP Analytics Cloud – Overview Guide

” Section 1.2

2024-04 (focus on analytics & planning

not AI/ML lakehouse).

5. SAP Help Portal

“SAP Datasphere – Product Documentation

” Section 2 ‘Data Integration & Semantic Layer

’ 2024-03 (integration focus; external engines required for ML).