At the SAP Sapphire user conference in Orlando this year, the great and the good from the ERP community gathered to hear Christian Klein, CEO of SAP SE explain where AI is going to be applied to the organization’s always meaty total software stack.
Klein of course started his keynote with a direct illustration of where AI agents are going to be applied in business and said that, today, 80% accuracy (which is roughly where many models deliver to) is just not good enough.
“Why do we face such huge challenges with AI in business, especially when LLMs are getting much better at tasks like writing code? It’s largely because a lot of the agentic services out there are trained on publicly available data; they’re not working with company-specific process knowledge that defines the brain of every company today – and I have to say that that ‘brain’ is still the ERP system inside any given enterprise,” said Klein.
Calling ERP the “trusted system of execution” in any company, the SAP leader used a working example of the SAP Knowledge Graph working alongside the company’s new SAP domain models to illustrate the organization’s AI-driven approach to smarter business processes. But because organizations need to work with one semantic data model, SAP systems are also carefully engineered to work with non-SAP models so that customers can be as relevant to their market as possible.
The SAP Agent Hub plays a central role here. This is a centralized platform for managing and deploying AI agents. It orchestrates autonomous workflows across enterprise applications, integrating AI LLMs with business data to automate complex tasks
CEO to CTO
SAP CTO Philip Herzig took over the mainstage keynote to walk attendees through the SAP Business AI Platform, where he says contextualized reasoning and governance AI are executed every day.
Herzig spoke at length about the SAP Process Consulting Agent from SAP Signavio, an AI-powered component designed to accelerate business transformation by acting as an autonomous, process-aware assistant. It works within the SAP Signavio Process Transformation Suite to analyze, recommend, and automate complex process tasks, acting as a “digital teammate” rather than just a tool.
Joule Studio is SAP’s new service for building, deploying and managing AI agents. Developers can build in their favorite language and framework, from Python to Claude Code to Cursor. Developers can then deploy to a managed runtime with no infrastructure management required. SAP users and partners can sign up for free design-time access for 12 months.
The SAP CTO quickly moved on to SAP Joule Studio 2.0, which offers agentic coding assistance and is a technology that is “open by principle” to allow developers to build inside the environment of their choice.
SAP Joule Studio 2.0
SAP Joule Studio 2.0 is a sophisticated low-code development environment designed to extend and customize the Joule Copilot service. It enables developers to build bespoke AI agents, refine generative AI behaviors and orchestrate complex business workflows. By surfacing enterprise data wherever needed, it ensures AI-driven tasks remain contextually aware and aligned with specific organizational needs.
A live demo of SAP Joule Studio 2.0 showcased how the software could be used to build a “pricing agent” i.e. an agentic service designed to price up a large organization’s total go-to-market offerings that might operate in anything from services industries to retail to manufacturing and so on.
Developers can use Custom Joule agents that understand context, orchestrate workflows and work with the total SAP platform to make use of agents capable of referencing existing business documents as trusted knowledge sources, while also using the elevation offered by Retrieval Augmented Generation (RAG).
Starting from a Product Requirements Document (PRD) SAP Joule Studio 2.0 is able to build a completed agent comparatively rapidly and produce a service (which may include multiple agents working together in an orchestrated way) to deliver skills that help organizations reduce pricing disputes across their organizational base.
SAP-RPT-1
Herzig went a little deeper and drilled into SAP-RPT-1, a specialized technical service within the SAP ecosystem designed to extract, analyze, and display specific organizational data. It typically facilitates automated reporting for financial or human resources modules, ensuring data consistency and streamlined decision-making through standardized, real-time business intelligence.
SAP moved to purchase Prior Labs in May of this year to bring a Tabular Foundation Models (TFM) skillset under the SAP umbrella. Prior Labs will continue to operate as an independent entity, with SAP committing to invest more than €1 billion over the next four years to scale it into a globally leading frontier AI lab for structured data.
According to Herzig and team, “Large language models struggle to make accurate predictions on structured business data because they have only a rudimentary understanding of tables, numbers and statistics. Unlike LLMs, TFMs are purpose-built for this type of data and can accurately predict business outcomes based on tabular data such as payment delays, supplier risks, upsell opportunities, customer churn risk and more.”
Other applications of AI in SAP were surfaced throughout this event.
AI for Revenue Management and Margin Performance is part of a new revenue management portfolio, which will help companies design, sell and profitably manage product, service or subscription models. It features a Billing Assistant, an agent for claims processing will automate rebate management across sales channels and manage unstructured claims, helping lower costs and increase settlement rates. New AI features will help improve margin insights by highlighting anomalies and settlement risks, allowing commercial and finance teams to intervene when profitability is at risk.
Take Your Partner…
SAP rounded out the Sapphire 2026 conference with a partner update. Given the massive presence that SAP takes and makes inside the Orange County Convention Center, it takes a little effort to navigate around the partner pavilion areas – it’s usually always worth the effort as the partners themselves are substantial firms (NTT DATA for instance is always there, not even with a huge stand) and they’re worth talking to.
In terms of “strategic” partnerships detailed this year, platform and suite partnerships include Anthropic, with Claude among the foundation models that SAP’s AI platform will be used to power Joule agents across HR, procurement and supply chain.
Amazon Web Services is bringing zero-copy data integration between SAP Business Data Cloud and Amazon Athena; Google Cloud and Microsoft are enabling bidirectional agent-to-agent interoperability between Joule and external agent frameworks.
Mistral AI and Cohere are delivering sovereign model options on SAP’s cloud infrastructure; n8n is providing visual AI workflow orchestration inside Joule Studio; and Nvidia is making sure its OpenShell (the company’s open source framework for building agentic AI applications) provides a trusted secure runtime for Joule Studio; and customer service company Parloa is bringing AI agents into SAP Service Cloud to handle customer interactions with full access to business data and service processes.
Joule in the Crown
SAP CEO Klein ended the keynote with a “live” discussion between himself and the company’s Joule AI service (illustrated with a jewel, rather than anything that might embody a unit of energy), allowing the agent to sum up the show itself.
“Let’s go! Christian. Okay, I need to be serious. I’m a new AI platform, together with our Autonomous Suite, giving you the ability to transform your company into an autonomous enterprise. Bringing the autonomous enterprise to life will help you to gain resiliency in times of uncertainty, when we sometimes don’t know what will happen tomorrow. But technology alone and plugging AI agents into your existing system landscape will drive zero value. Moving to the autonomous enterprise requires serious change management. Adoption of AI goes hand in hand with business process redesign and end-user enablement,” said the service.
All of which allowed the company to table one final AI offering, the new Rise and Grow with SAP service. This sees SAP architects, AI developers, industry consultants, as well as Line of Business (LOB) consultants all come to customers’ premises on-site to drive change management and AI adoption.
