A steady stream of ERP innovations from SAP

SAP Cloud ERP, Joule and Agentic AI, quantum computing

Over the past 24 months, the Walldorf-based software group SAP, under the leadership of CEO Christian Klein, has performed a breathless and panic-driven U-turn in order not to lose touch with the fast-paced age of artificial intelligence. Driven by the existential fear of the financial markets that the classic ERP business could become obsolete due to autonomous AI agents, SAP buried the North Star architecture that had only recently been propagated and proclaimed the new target image of the „Autonomous Enterprise“.

A critical look at the innovation and acquisition strategy reveals that SAP is heavily reliant on inorganic growth through acquisitions worth billions and an extensive cooperation network due to a lack of its own fundamental AI core competencies. In order to create the urgently needed database for future AI agents, SAP acquired the data specialists Dremio and Reltio in recent months, whose technologies are intended to help query and harmonize data from a wide variety of systems without prior copying.

At the same time, the Group bought the Freiburg-based start-up Prior Labs to build a world-leading AI lab and develop spreadsheet models for business applications. This massive purchasing spree was supplemented by the more recent acquisitions of LeanIX for enterprise architecture management, WalkMe to provide a digital adoption platform and SmartRecruiters to expand its own HR management in the SuccessFactors environment with AI-supported recruiting.

Missing LLM but n8n, Parloa and SAP Joule Studio

As SAP does not develop its own basic models (large language models) such as OpenAI, the Group instead integrates technology from third-party providers and strategically invests in agile start-ups. Recent investments in the millions were made in the Berlin-based automation start-up n8n, whose platform will be integrated into the SAP development environment Joule Studio in future to make it much easier for customers, consultants and developers to create their own AI agents.

SAP also deepened its partnership with the Berlin-based tech company Parloa, which specializes in AI agents for contact centers, in order to strengthen its own customer experience portfolio and automate service requests to a large extent.

On a global level, SAP entered into far-reaching strategic alliances with Nvidia to develop complex reasoning agents based on the Llama-Nemotron models for autonomous problem solving. Llama-Nemotron models are a family of open AI models developed by Nvidia that have been optimized for agent-based workflows and advanced reasoning. They are characterized by open source availability and innovative features such as a dynamic reasoning toggle.

Business Technology Platform (SAP BTP) and Business Data Complexity (SAP BDC)

There is also cooperation with Google, Meta, Mistral AI, Microsoft and Perplexity to make their language models available to existing customers via the Generative AI Hub of the Business Technology Platform (SAP BTP). In the data environment, fundamental partnerships were also established with Databricks and Snowflake in order to anchor the new SAP Business Data Cloud (BDC) in the market as an open data platform that is strictly controlled by SAP.

In terms of products and innovations, SAP is bundling these new ERP functions in the SAP Business AI Platform, which brings together the technical pillars of BTP, BDC and tools such as Signavio and LeanIX in one environment. Building on this, SAP promises the „SAP Autonomous Suite“, which will autonomously handle complex business processes - from procurement to quarterly closing in the finance department - with over 50 domain-specific Joule assistants and an armada of over 200 highly specialized AI agents.

Graph theory and computer science: SAP Knowledge Graph

An essential architectural building block that SAP is launching is the SAP Knowledge Graph. This knowledge graph is intended to semantically translate the deep, often cryptic SAP database structures for the algorithms in order to avoid dangerous AI hallucinations and provide the language models with the indispensable business context.

However, analysts and the user association DSAG warn that this autonomous target image is currently primarily an effective advertising ploy, as the well-known cloud solutions such as Ariba or the Business Network continue to operate in the engine room of the Autonomous Suite and have merely been enriched with new AI features. This strategy is also flanked by a highly controversial API policy with which SAP rigorously regulates data access from external AI services. Experts see this as a ruthless attempt to restrict the digital sovereignty of users and force them deep into its own lucrative AI ecosystem of SAP BDC and SAP Datasphere.

Joule Work and Autonomous Enterprise

However, the most radical paradigm shift for existing SAP customers is manifested in Joule Work, the Walldorf-based company's newly unveiled central user interface. With Joule Work, traditional forms, menu trees and historical transaction codes are finally a thing of the past; instead, users simply describe their intentions or tasks in natural language via text input, while the AI assistant Joule proactively and independently triggers and executes the necessary work steps and specialized agents in the background.

For existing SAP customers, this development is a double-edged sword: on the one hand, it promises huge leaps in productivity and a simplification of the often complex SAP operation; on the other hand, there is the threat of a blatant loss of control if autonomous agents make business-critical decisions in a black box. In addition, insiders from the SAP community warn that marketing is miles ahead of actual technological maturity, as Joule often still provides incorrect answers in the tough day-to-day business environment or fails due to the complex realities of historically grown SAP databases.

SAP licensing strategy, AI metrics and monetization

However, the greatest financial risk for the user is hidden in the monetization strategy behind these innovations: If AI agents take over manual knowledge work in the future, this will inevitably devalue SAP's traditional, user-based software licenses. The Group is therefore moving full steam ahead towards consumption-based pay-per-AI usage models. These new metrics harbour the massive risk that the much-vaunted Autonomous Enterprise will mutate into an incalculable, permanent cost trap for existing SAP customers, in which every process automated by AI is expensively billed and the digital sovereignty of users is sacrificed on the altar of maximizing margins in Walldorf.

QUTAC: Quantum computing and the future of ERP

The Walldorf-based software group SAP is increasingly adorning itself with the futuristic concept of quantum computing as part of its innovation promises, but a closer analysis often reveals this commitment to be a strategic smokescreen to distract attention from its own shortcomings in the traditional ERP and AI business.

Although SAP has founded the Quantum Technology and Application Consortium (QUTAC) together with well-known German companies such as BASF, BMW, Bosch and Siemens in order to create the technical basis for industrial usability in Europe, Walldorf's actual commitment remains nebulous.

When asked by E-3 Magazine what concrete plans the Group is pursuing within this consortium, SAP remained eloquently silent, which reinforces the suspicion that the global ERP market leader is currently merely taking on the role of a silent follower in this highly complex topic instead of building up its own real core competencies. From a critical perspective, these excursions into the still immature sphere of quantum computing primarily seem like a naïve waste of resources without any tangible ERP added value, especially as SAP has not even completed the fundamental homework of integrating artificial intelligence into its own S/4 core to the satisfaction of the community.

Nevertheless, an existing SAP customer should understand the theoretical architectural and business objective of linking ERP systems and quantum computing in the future: While classic computers and supercomputers process information sequentially in bits, quantum computers operate with so-called qubits that can assume an infinite number of states simultaneously, which in theory enables the solution of highly complex mathematical problems that today's IT architectures inevitably fail to solve.

Quantum algorithms and supply chain management

For enterprise resource planning, this means a paradigm shift in the management of extreme complexity, which today simply overwhelms conventional ERP systems. The focus here is on huge optimization tasks within logistics and the global supply chain. In future, quantum algorithms should be able to calculate the optimal warehousing, the ideal warehouse positioning, the most efficient route planning, taking into account countless volatile environmental variables, as well as profound risk minimization in a fraction of a second. In addition, the transformative technology is aimed at energy optimization in factories and office buildings, complex simulations in marketing and revolutionary breakthroughs in cryptography to ensure secure, encrypted communication and data storage in the ERP system of the future.

At present, users are not actually benefiting at all from SAP's day-to-day operations. The technology is still in its infancy worldwide and analysts warn against being blinded by the bold marketing buzzword that SAP is using to present investors on the financial markets with a cloudy innovation story beyond traditional cloud computing. If more than 50 percent of companies see quantum computers as a decisive competitive factor of the future, then this is the result of the correct realization that highly complex supply chain calculations or the simulation of industrial and chemical processes will only be solvable with this computing power.

For SAP users, this means an educational mission to strategically keep the development of quantum computing on the radar in order to be prepared for future, decentralized real-time operating systems, but not to be distracted by the Walldorf PR machinery for the foreseeable future. A confident IT decision-maker must recognize that although the target image of an intelligent company (autonomous enterprise) will require the computing power of quantum chips in the long term, the current priority must be to clean up the company's own historical custom code legacy, manage the upcoming S/4 migration and use current, market-ready AI tools in a beneficial way before valuable budgets flow into SAP's half-baked quantum visions.

SAP CEO Christian Klein is hard-working and ambitious. However, the global SAP community lacks a sound and continuous strategy. Every month, a new ERP idea is born, a new AI product is presented and an AI start-up is acquired. To keep up with Christian Klein, an existing SAP customer needs a lot of time, personnel and capital. The vast majority of ERP users do not have the resources to follow SAP into the AI and quantum future: The time to move fast is now, isn't it?