Agentic AI Needs Open Data, Not Bigger Monoliths
Enterprise agentic AI is the use of autonomous, goal-driven software agents that operate across finance, logistics, and other functions by querying, reasoning over, and acting on data from many systems in real time, rather than being confined to one application stack or a single vendor’s database. SAP’s Dremio acquisition and Nokia’s RISE with SAP agreement both point to the same conclusion: the future of enterprise AI will not live inside isolated ERP instances, but on hyperscale cloud platforms that treat data integration as a first-class design problem. By completing the SAP Dremio acquisition, SAP is no longer betting that tighter lock-in will win; it is betting that enterprises will reward vendors who can orchestrate heterogeneous data and AI workloads at scale. That is a strategic shift away from monolithic software and toward cloud-native data fabrics.
Dremio: SAP’s Bet on Enterprise Data Integration for Agentic AI
SAP has completed its acquisition of Dremio, an open, high-performance data lakehouse platform that directly targets the enterprise data integration bottleneck. SAP says the deal “accelerates agentic AI and expands customers’ ability to combine SAP and non-SAP data to run analytical and AI workloads in real time, with no data movement or conversion necessary, and with vastly improved economics for enterprise analytics”. In plain terms, Dremio’s data virtualization means SAP no longer insists that all meaningful analytics must sit inside S/4HANA; instead, SAP is accepting that valuable data lives in many systems, and the AI layer must see all of it. This breaks the old incentive to centralize everything in one stack. It also removes a major technical barrier to agentic AI deployment, because autonomous agents can query across SAP and non-SAP data without waiting for nightly ETL jobs or complex duplication pipelines.
Nokia Shows Cloud ERP Transformation Is Now an AI Architecture Choice
Nokia’s multi-year agreement with SAP to transform its ERP through RISE with SAP, with its SAP S/4HANA environment hosted on Microsoft Azure, is more than a migration story. Nokia is consolidating multiple ERP systems into a unified S/4HANA landscape for finance and key logistics, while SAP operates and manages the environment in the cloud on Azure’s hyperscale infrastructure. The point is not only clean-core ERP; it is where AI will run. Large SAP customers now treat hyperscaler choice as a strategic decision, because the cloud platform shapes performance, security, resilience, data strategy, and how fast AI-enabled capabilities can be adopted. By standardizing processes and reducing customization sprawl, Nokia is building a backbone that can support continuous embedded AI and cloud innovation rather than frozen on-premises infrastructure. In this model, AI is enabled at the hyperscaler layer, close to elastic compute and cross-system data, not just inside the ERP database.
Hyperscalers Become the AI Enablement Layer, Not Mere Hosting
The Nokia agreement clarifies how hyperscalers have moved from background infrastructure to the AI enablement layer. Azure is not only hosting S/4HANA; it provides the global scale, security, and performance needed to support Nokia’s most business-critical workloads while SAP delivers the managed cloud operating framework. Nokia ends up with a three-party model: SAP for ERP and cloud operations, Microsoft for hyperscale infrastructure, and Nokia for business transformation and AI-enabled processes. This architecture aligns with the SAP Dremio acquisition. If AI agents must query SAP and non-SAP data in real time with no data movement, the logical place to run them is a hyperscale cloud that already aggregates storage, analytics, and security controls. Enterprise AI ambitions are therefore shifting from monolithic on-premises stacks to cloud-native designs where data integration and AI orchestration happen at scale, across workloads and vendors rather than inside one application silo.
From Locked-In Stacks to Open Fabrics: What CIOs Should Do Next
SAP’s embrace of Dremio and Nokia’s RISE with SAP path on Azure should be read as a warning to CIOs still chasing a single-stack dream. Agentic AI deployment will falter if data remains trapped in fragmented ERP environments, inconsistent master data, and brittle customizations. Cloud ERP transformation is now inseparable from AI strategy: clean-core discipline in S/4HANA, enterprise data integration across SAP and non-SAP systems, and a hyperscaler layer capable of real-time analytics are no longer optional. The practical takeaway is clear. Stop designing AI pilots around one application; start designing around the data fabric and the hyperscaler. Prioritize architectures that allow agents to query across systems without ETL delays, keep customization in check, and treat hyperscaler AI enablement as a core part of ERP planning. Enterprises that make this shift will move faster, spend less time fighting integration debt, and give AI agents the context they need to be useful.
