Key Takeaways
⇨ SAP Industry AI builds industry-specific processes, rules, data and operational context into AI applications and agents.
⇨ The SAP Business AI Platform provides the underlying data, models and governance, while Industry AI adds industry-specific specialisation.
⇨ SAP is using AI agents and forward-deployed engineers to move from customer-specific solutions towards repeatable, increasingly autonomous workflows.
SAP has introduced an Industry AI portfolio aimed at enterprise challenges where better AI-generated recommendations alone are not enough.
SAP points to challenges such as coordinating thousands of field technicians following an energy-grid outage, keeping production lines running during supply-chain disruption, and automating pharmaceutical batch release without compromising quality or regulatory controls.
Dominik Metzger, President of Industry AI at SAP, summed up the challenge plainly: “These problems are incredibly hard to solve.”
What Is SAP Industry AI?
SAP describes Industry AI as its approach to building AI around the processes, rules, data and exceptions that define individual industries. Rather than starting with a general-purpose model and adapting it afterwards, SAP aims to build industry context into the applications and agents carrying out the work.
The ambition is to move beyond AI that summarises information or recommends a next step. Industry AI is intended to support systems that can increasingly coordinate decisions and actions across complex business processes.
How Is SAP Industry AI Different from SAP Business AI?
SAP positions the Business AI Platform as the technical foundation beneath Industry AI. It provides the models, data, integration and governance capabilities needed to build and run AI across the SAP landscape.
Industry AI adds industry-specific specialisation. Joule, meanwhile, acts as the engagement layer through which users interact with assistants and agents.
Put simply, the Business AI Platform provides the foundation, Joule provides the interface, and Industry AI adds the sector-specific context needed for more specialised execution.
How Does SAP Industry AI Work?
SAP Business Data Cloud provides contextual enterprise data, while SAP Domain Models draw on the company’s institutional knowledge to help AI systems understand business concepts and relationships.
SAP AI Agent Hub provides the management layer. Businesses can set boundaries around how agents, applications, large language models and MCP servers operate, while measuring results against defined goals.
Industry AI brings these components together within specific operational scenarios, where agents can use current business data and industry knowledge to work across processes that may span several applications, teams and business functions.
What Industry AI Applications Is SAP Building?
SAP has identified seven priority Industry AI domains: Asset Management, Commodity Management, Adaptive Production, Regulated Manufacturing, Revenue Growth Management, Unified Commerce and Project Delivery.
The range shows that SAP is applying this approach well beyond manufacturing. Asset Management targets reliability, uptime, safety and compliance in asset-intensive businesses. Adaptive Production supports complex configure-to-order and engineer-to-order workflows. Unified Commerce brings together merchandising, planning, marketing, shopping and fulfilment.
Each domain focuses on a business process that crosses traditional application boundaries and applies AI to the decisions and actions required to keep it moving.
Why Is SAP Using Forward-Deployed Engineers?
SAP’s forward-deployed engineering model embeds specialists, including data scientists and AI engineers, directly with customers to tackle problems that do not yet have packaged solutions.
SAP says it is beginning with selected ECC and Private Cloud customers, using these engagements to test and refine solutions in real operating environments. Where an approach proves repeatable, SAP intends to turn it into standardised systems of agents that can be deployed more broadly.
That makes forward-deployed engineering more than simply a services model. It is also part of how SAP plans to identify which Industry AI use cases are mature enough to become scalable products.
How Does Industry AI Fit Into SAP’s Autonomous Enterprise Strategy?
SAP describes the Autonomous Enterprise as the broader destination: people set direction, assistants coordinate and agents execute, with actions governed and measured against business outcomes.
Industry AI gives that model industry-specific depth. The idea is that an agent working in utilities, manufacturing, retail or another sector should understand more than the general mechanics of a workflow. It should also understand the industry rules and operational context that determine what a valid decision looks like.
That is the broader shift SAP is pursuing: moving from enterprise software that primarily records activity and supports decision-making towards systems that can increasingly carry out parts of the work themselves.
What Should SAP Customers Watch as AI Moves Towards Autonomous Execution?
The first question is how much authority customers are prepared to give agents. Recommending an action, initiating a workflow and independently executing a consequential business decision require very different levels of oversight.
Governance therefore becomes part of the buying decision. Customers will need to understand how agent permissions are defined, how actions are monitored and audited, when humans remain in the loop, and how an incorrect action can be stopped or reversed.
The other question is scalability. SAP’s forward-deployed model depends on turning customer-specific work into repeatable products. How quickly SAP can make that transition will help determine whether Industry AI develops into a broadly deployable software portfolio or remains concentrated in highly customised engagements.
What This Means for ERP Insiders
Decision authority is moving closer to AI agents. Customers should build testing, human oversight, escalation routes and rollback controls into deployment plans as agents take on greater operational responsibility.
Governance becomes a procurement checkpoint. IT, security and risk teams should evaluate how agent permissions, actions and audit trails are controlled before expanding autonomous execution.
Early Industry AI deployments may look different from conventional software roll-outs. SAP is developing some of these capabilities directly with ECC and Private Cloud customers before standardising them for wider deployment.
