Intelligent error analysis for SAP planning processes using generative AI
IPEC
Project description
Decision-making processes. Errors in underlying data, models, or workflows can significantly compromise the quality of decisions and results. Despite rapid advances in generative AI and large language models (LLMs), their potential for in-depth analysis and correction of such errors in SAP-based systems has so far been utilized only to a limited extent.
The IPEC project addresses this challenge. The goal is to develop a solid foundation for the future use of AI- and LLM-based methods for intelligent error analysis and correction in the client’s planning processes.
The first phase of the project focuses on systematically identifying existing challenges, developing a shared understanding of the problem, and structurally decomposing the problem statement. To this end, requirements will be gathered, relevant use cases identified and prioritized, and potential development paths evaluated. The results will form the basis for subsequent project phases, in which suitable methods will be developed, prototyped, and gradually refined into a scalable Minimum Viable Product (MVP).
By identifying suitable use cases at an early stage, the project lays the groundwork for the targeted use of modern LLM and GenAI technologies in complex analysis and decision-making processes within planning systems. At the same time, the scope, roadmap, and technical priorities for the subsequent development phases are defined.
Research contribution
fortiss is responsible for the scientific analysis of the problem and the methodological design of the approach. This includes requirements workshops, a detailed examination of existing challenges, the structured decomposition of the problem, and the prioritization of relevant use cases.
In addition, fortiss develops a well-founded decision-making document that serves as the functional and technical foundation for the subsequent project phases. The work is carried out according to an agile process model to validate requirements early on and set the course for the successful development of AI-based solutions.
Project duration
16.02.2026 - 31.12.2026
Contact
Project partner
