Ericsson scales up AI with datafabric and SAP

Ericsson is deploying SAP Business Data Cloud and the Joule AI copilot across its global organisation. This will replace the fragmented AI experiments with a unified data architecture designed to scale. SAP announced the partnership at SAP Sapphire 2026.

The Swedish telecoms infrastructure company operates in 180 countries and carries more than 40% of global mobile data traffic. As AI becomes central to both its technology strategy and day-to-day operations, Ericsson has opted for a federated data architecture. In which data remains in source systems while semantics, governance and lifecycle policies are managed centrally.

The approach avoids unnecessary data duplication. And ensures that business definitions, for revenue, market structures and access rights, are applied consistently across SAP environments and beyond.

From AI experimentation to enterprise scale

That consistency is what makes AI tractable at scale. More than 85,000 employees now work with the integrated version of Joule, SAP’s generative AI copilot. Ericsson credits executive sponsorship and clear governance frameworks with making that rollout possible without losing control of the underlying data.

The transformation runs on two parallel tracks. The first is modernisation: migrating to RISE with SAP, building new functionality on SAP Business Technology Platform rather than inside the ERP core. And maintaining a clean-core approach that keeps the system extensible without accumulating technical debt.

The second track, which Ericsson calls “innovate and transform”, uses data and AI to generate operational impact: faster decision-making, more efficient processes and new value creation models.

SAP and Ericsson are also co-developing new AI capabilities. One example is an intelligent feature within SAP SuccessFactors. This automatically suggests relevant objectives for employees based on their role and the company’s strategic goals. Thus, reducing administrative overhead while improving alignment.

Esra Norell, Vice President Customer Experience Enterprise IT at Ericsson, framed the move in direct terms: “Once you start scaling AI, it stops being an AI question. It becomes a data question. That is why we invested early in a datafabric.

“With SAP Business Data Cloud, we can centrally define how we understand data, such as revenue, market structures and access rights, and then apply that meaning consistently across the organisation. That is what makes it possible to scale AI in a reliable, repeatable way and deliver real business value.”

Manos Raptopoulos, Chief Revenue Officer EMEA, MEE and APAC at SAP, added: “Ericsson’s approach shows how leading companies move from AI experimentation to actual enterprise deployment, focusing on data, governance and business context. Together, we are helping organisations deploy AI at scale and get the most out of the technology.”

Enterprise Times: What does this mean?

Ericsson’s deployment is notable less for its scale than for its architecture. The decision to keep data in source systems and govern semantics centrally, rather than pulling everything into a single data lake, reflects a hard-won lesson from earlier enterprise AI rollouts: bad data governance does not disappear when you add an AI layer on top. It amplifies.

The federated model Ericsson has chosen maps well onto the realities of a multinational operating across 180 countries with multiple regulatory environments and legacy system landscapes.

For Ericsson itself, the stakes are high. The company is not running a pilot. With more than 85,000 employees already using Joule, this is an operational commitment. With the two-track approach, modernising the ERP core while simultaneously running AI innovation initiatives, is a deliberate hedge: keep the business running while building the capabilities that will define competitiveness in the next cycle.

Beyond the pilot: what Ericsson’s approach signals

The emphasis on measurable outcomes, faster decision-making, process efficiency, new revenue models, also signals that Ericsson’s leadership is under pressure to show returns on its AI investments, not just adoption numbers. Shareholders will be watching whether the datafabric investment produces the productivity gains and new digital revenue models the company has put on the record.

For SAP, the partnership is a validation of the Business Data Cloud proposition at the enterprise tier. Ericsson is exactly the kind of reference customer SAP needs: a global, complex, multi-system operator that has chosen the federated approach over competing architectures. Every CIO considering a similar move will look at how this plays out.

The co-development of new Joule capabilities, including the SuccessFactors goal-setting feature, points to a deepening of the relationship beyond standard deployment. Which has implications for SAP’s product roadmap and its positioning against Microsoft Copilot and other AI platform competitors.

The clean-core discipline Ericsson has followed, building new functionality on BTP rather than in the ERP core itself, also reinforces a message SAP has been trying to land with its customer base for years. Seeing a company of this complexity execute on it, and then layer AI on top, makes the argument more concrete than any analyst report can.

The implications extend beyond the two companies. Ericsson supplies network infrastructure to operators in 180 countries. If the company’s AI-driven decision-making and operational improvements translate into faster or more reliable delivery for those operators, the effects ripple outward across an industry. That is itself under pressure to justify its own infrastructure investments. In an era of margin compression and accelerating technological change.

The post Ericsson scales up AI with datafabric and SAP appeared first on Enterprise Times.