KI-S/4 kick-off
Beyond the hype surrounding artificial intelligence, it is becoming increasingly apparent that the targeted use of AI technology enables tangible increases in efficiency in complex ERP projects. Particularly in transformation projects such as S/4 migrations, AI can significantly reduce the burden on existing SAP customers and speed up processes. This leads to measurable time and cost savings as well as an increase in quality in critical project phases.
A prime example of this potential of AI is the recently completed project at an Ososoft customer. Here, the advantages of AI
The use of S/4 Hana in a complex S/4 Hana transformation project can be demonstrated under real conditions.
Mastering complexity: S/4 migration of a global construction company. Our client, a construction company with an annual turnover of over ten billion euros and more than 30,000 employees in over ten countries, was facing typical problems in the midst of its S/4 transformation: insufficiently documented changes and processes, a lack of training documents, a lack of test and regression test cases. These deficits were the result of a shortage of specialists and the double burden of day-to-day business and project tasks on the specialist department.
During this critical project phase, it became clear that innovative solutions were needed to ensure the success of the project. We recommended the use of artificial intelligence to optimize the most resource-intensive tasks. The focus was on the accelerated, AI-supported creation of tests, documentation and training documents, while at the same time reducing the workload of the specialist department.
Intelligent automation
Our approach was based on three pillars: firstly, efficient data collection through focused process recordings; secondly, AI-supported processing to generate test cases, documentation and training materials; and thirdly, multilingualism. Through short, focused interviews (on average 30 minutes per process), we collected detailed data on process flows and modifications. We used this data to generate precise AI-supported test cases, create technical and user-centered documentation and develop tailored training materials in six languages.
AI-supported best practice
The key to success lay in the combination of efficient data collection and AI-supported processing. Instead of time-consuming workshops, we implemented a lean, structured process to gather information. Through short, targeted interviews with key people, we were able to collect detailed data on process flows, technical modifications and best practices. With an average duration of just 30 minutes per process, we were able to reduce the burden on the specialist department to a minimum while building a robust and consistent database for our AI systems.
We use the data created in this way to fundamentally optimize three core areas. In the area of testing, we developed an AI module that analyzes requirements, selects relevant processes and generates precise test cases. These are transferred directly to the automation software used, which significantly speeds up the testing process.
Documentation and modifications
For the documentation, we first integrated the company-specific authoring guidelines into our AI system. This allowed us to create two types of documentation: technical documentation, which precisely captures all changes, modifications and customizing aspects of a process, and user-centric documentation, which focuses on the process flow and contains detailed background information and department-specific explanations.
For the training area, our system used the documentation created to generate customized training materials. It automatically created presentations and videos that could be tailored to different user groups. The system took into account factors such as granularity, prior knowledge and complexity in order to control the focus of the respective training courses and thus meet the needs of the respective target group.
The integrated multilingualism proved to be a particular advantage: all generated materials - from test cases and documentation to training materials - could be created in six different languages according to the requirements. This enabled a smooth and accelerated international roll-out of the S/4 Hana system without any additional work for the specialist department or external translation service providers.
This approach made it possible to create high-quality, tailor-made materials with minimal impact on the specialist department. The use of AI not only acted as an efficiency booster, but also as a real catalyst that fundamentally changed the project dynamics and set new standards in terms of time-to-value.
Measurable project success with AI
Overall, our AI-supported approach led to measurable improvements in the affected project areas and made a significant contribution to meeting critical milestones. In just three days, we were able to generate materials that would have taken around 180 man-days to create conventionally. This enormous time saving made it possible to meet critical milestones despite the tight resource situation and even carry out additional quality assurance measures.
An initial challenge was the skepticism of employees towards AI, especially the fear of being replaced. Through workshops and demonstrations, we were able to show that AI relieves employees of repetitive tasks, not replaces them. This was essential for the initial acceptance of the project and the support of the specialist departments.
There were significant improvements in each of the core areas of the project: In testing, the effort required to create test cases was reduced by 65%, while test coverage increased by an impressive 220%, which led to a significant increase in project security.
Training material
The creation of technical and procedural documentation was accelerated by over 80 percent, resulting in more comprehensive and up-to-date documentation. A time saving of 57% was achieved in the development of training materials, which enabled employees to familiarize themselves with the new system more quickly and effectively. A key advantage of the approach was the low time burden on the specialist department: with an average of just 30 minutes of work per process, the technical experts were able to contribute their knowledge efficiently without being distracted from their core tasks - overall, the specialist department was able to save around 120 budgeted man-days through the targeted use of AI. This contributed significantly to the acceptance of the project and enabled the department to focus on critical aspects of the transformation.
Interestingly, the quality of the materials generated even exceeded our own expectations. After initial difficulties, they were characterized by a high level of consistency, depth of detail and target group orientation - an impression that was confirmed by positive feedback from various areas of the company. Incorrect information and hallucinations were almost completely eliminated thanks to the data provided.
Conclusion
The use of AI in this S/4 Hana transformation project has clearly shown that artificial intelligence is not just hype, but can definitely help to master complex challenges in digital transformations. The time and cost savings achieved as well as the measurable increase in quality in almost all project phases show the enormous potential of this technology.
The fact that the use of AI was not at the expense of employees is particularly noteworthy. On the contrary: the AI took over repetitive tasks, relieving the burden on employees and allowing them to concentrate on more value-adding activities. This led to higher employee satisfaction and ultimately to a successful project completion.
The insights gained from this project serve as a valuable basis for future transformation projects. The use of AI as a strategic catalyst will play an increasingly important role and help companies to complete their digital transformation more efficiently and quickly.