What the study found
The study found that agentic AI, a form of AI that can coordinate multiple steps autonomously, can generate dynamic decision tables from business process model and notation (BPMN) models. The authors report that this approach can identify decision points and produce optimized decision model and notation (DMN) tables.
Why the authors say this matters
The authors say this matters because manual creation of DMN decision tables in BPMN environments can introduce human error, inconsistencies, and cognitive bias. They conclude that the approach could support intelligent, adaptive decision support systems in mission-critical environments and autonomous decision modeling that can adapt to changing business requirements.
What the researchers tested
The researchers developed an AI-based system to generate decision tables from BPMN models. The system used large language models inside an agentic AI framework, with agents for BPMN analysis, decision extraction, rule generation, and validation, coordinated through a ReAct engine and retrieval-augmented generation (RAG).
What worked and what didn't
In the experimental evaluation of critical applications, the system reportedly suggested decision tables with values that humans might not intuitively identify. The findings indicate that it can turn ambiguous process paths into more precise decisions and identify non-obvious decision criteria and threshold parameters, which the abstract describes as improving process automation.
What to keep in mind
The abstract does not provide detailed performance metrics, comparison conditions, or failure cases. It also does not describe limitations beyond the general focus on critical applications.
Key points
- Agentic AI was used to generate dynamic decision tables from BPMN models.
- The system was designed to identify decision points and optimize DMN tables.
- The authors say manual DMN creation can introduce human error, inconsistencies, and cognitive bias.
- The evaluation found suggestions that humans might not intuitively identify.
- The abstract says the system identified non-obvious decision criteria and threshold parameters.
Disclosure
- Research title:
- Agentic AI generated DMN tables from BPMN models
- Authors:
- Sourour Meddeb, Selma Batti, Habib Fathallah
- Institutions:
- University of Carthage, University of Carthage, University of Carthage
- Publication date:
- 2026-01-21
- OpenAlex record:
- View
Get the weekly research newsletter
Stay current with scholarly research without reading academic papers — one filtered digest, every Friday.