Inside SAP

April 2026 feels different from the SAP AI months that preceded it. Not because of any single announcement, but because the infrastructure, data, and application layers are converging in ways that make the overall direction harder to dismiss. After nearly three decades of watching SAP manage the gap between announcement and delivery, I am cautiously more optimistic than I was six months ago — with the emphasis on cautiously. The infrastructure story is particularly encouraging. SAP's deepened NVIDIA partnership, with NeMo model training now available through AI Core, suggests SAP is serious about supporting more sophisticated enterprise AI workloads. Equally important is SAP Business Data Cloud reaching general availability with Snowflake zero-copy integration — this addresses one of the biggest practical barriers to AI adoption: data accessibility across siloed systems without the usual ETL headaches. SAP's agreement to acquire Reltio further reinforces this data focus, bringing master data management capabilities that make both SAP and non-SAP data more AI-ready. On the application side, SAP's claim of 400+ Business AI use cases with 40 Joule agents shows they are moving from prototype territory into scaled productization. The ABAP-1 foundation model trained on 250 million lines of code could be genuinely useful for modernization efforts, though customers will need to test it against their real development patterns before declaring victory. Sapphire 2026 in Orlando next month becomes the logical proving ground. The Celonis lawsuit introduces the one discordant note. While legal battles between ecosystem players are rarely customer emergencies, they do highlight the growing tension around data access and platform control. Customers should use this moment to assess their process intelligence dependencies and ensure they maintain optionality regardless of how the vendor relationship evolves. This convergence of infrastructure readiness, data platform maturity, and application-level AI suggests SAP AI may finally be entering a phase where operational value becomes measurable rather than merely promised. The real test will be whether customers can deploy these capabilities without the usual process redesign, data cleansing, and change management prerequisites that have historically slowed SAP innovation adoption.