LLM support for non-expert enterprise modeling

Formalizing Model Selection in LLMOps: A Systematic UML-Based Process Model

The proposed model aims to support the formalization of model selection processes, improve decision-making, and enhance the traceability and transparency of LLMOps practices and forms part of a broader research effort toward the formalization of the entire LLMOps life cycle.

A tool selection mechanism for LLM-based model management agents

In Model-Based Engineering (MBE), practitioners frequently have to choose appropriate tools from many different options. LLM-based agents are software components that depend on Large Language Models (LLMs) to autonomously select and apply software tools to perform specific tasks. Although LLMs have already been used in the MBE context, LLM-based agents to assist users of MBE tools remain underexplored. This is particularly true in industrial environments where only medium-sized on-premise LLMs can be considered due to policies related to security or data privacy. To investigate the potential of LLM-based agents for MBE, we study model management tools for transformation and editing. Currently, off-the-shelf agents such as Microsoft Copilot can invoke model management tools when the task is explicitly described. However, these agents struggle to select the correct transformation or operation when they only have limited contextual information, especially when coupled with medium-sized LLMs. To overcome this, we propose an approach based on complementary mechanisms. First, we provide a server and associated LLM-based agent with dedicated tools for each transformation available on this server. We also provide a similar server and agent for model editing operations. Then, to enable the two agents to efficiently select transformations and operations (respectively), we rely on a tool retrieval technique based on a tool relevance score computed by an LLM. We evaluate these agents using generated model management datasets that we contribute to the community. The obtained results show that our LLM-based agents select more accurately modeling tools to perform user instructions than off-the-shelf agents.

Integrating Human-factors into Enterprise Architecture Leveraging Ontologies and Metamodels

This paper builds on the author’s previous work regarding domain-specific ontologies (DSO) and its importance in the human-factors integration (HFI) space. Explicit term definitions captured by a DSO allow the HFI vocabulary to be mapped into a model-based enterprise architecture (MBEA). Integrating this terminology into the overall MBEA provides insight into the role that individuals play by considering personnel as a critical system component. Often considered external actors, human resources are typically not accounted for in the original solution design. However, MBEA promises to reverse this trend by implementing the Unified Architecture Framework (UAF). The UAF is composed of various domains and their aspects and is meant to graphically illustrate enterprise concepts such as strategy, operations, resources, personnel, and services in a digital environment. Capturing HFI information in a model improves the traceability of person(s) and organizational concerns, responsibilities, and competencies to highlight gaps that must be addressed. The incorporation of the HFI DSO into an MBEA enhances communication between disciplines and provides transparency for stakeholders. This research demonstrates the feasibility of constructing a DSO based on an HFI body of knowledge; leveraging the Web Ontology Language (OWL), the subject-predicate-object (SPO) approach, and the Protégé ontology editor. It also shows that by importing the OWL file into a concept model, understanding HFI terms facilitates MBEA while maintaining personnel as a critical part of a successful organization. This research identifies areas for improvement of the UAF domain-specific modeling language (DSML) to ensure that it adequately addresses HFI concerns by mapping like-terms.

Automated Business Process Model Generation With Large Language Models: A Systematic Literature Review

The automated generation of business process models from natural language descriptions has recently attracted growing attention at the intersection of Business Process Management (BPM), Natural Language Processing (NLP), and Large Language Models (LLMs). This paper presents a Systematic Literature Review (SLR) on the current state of research in this emerging field. Following the guidelines of Kitchenham et al. and the PRISMA framework, 29 studies published between January 1st, 2023 and March 10th, 2026 were identified, selected, and analyzed. The review addresses two research questions focusing on the applied methodological approaches, the used LLMs, the employed modeling languages, as well as the evaluation strategies, challenges, and limitations reported in the literature. The results show a clear shift from traditional NLP-based techniques toward LLM-only and hybrid approaches. OpenAI’s GPT family, especially GPT-4 and its variants, dominates the field, while BPMN is by far the most frequently used target process modeling language. Furthermore, existing studies evaluate automated process model generation primarily through output-focused methods, such as quantitative metrics, expert reviews, and comparisons with alternative or human-created process models. At the same time, the reviewed studies reveal important challenges, including the continued need for human involvement and the output quality. Overall, current approaches show strong potential, but they still act more as intelligent assistants than as fully autonomous process modelers.

Automated Generation of SysML Activity Diagrams from Industrial Requirements Using LLMs

A model-driven knowledge graph treatment for process-centric cyber threat mitigation

Cybersecurity analysis touches on diverse interdisciplinary aspects, both technical and business-oriented, and diverse stakeholders from different backgrounds must be supported with a knowledge capture environment that operationalizes the “secure by design” principles while also ensuring semantic traceability of the design decisions, to enable AI-based analysis. This paper introduces a domain-specific modeling method supported by a process-centric visual language (a BPMN extension) to enable a knowledge graph treatment for designing cyber threats mitigations in tandem with business process engineering activities and associated data flows. The treatment is intended to be leveraged by LLM (large language model) services, therefore experimentation leans towards AI integration. The domain-specific modeling language (DSML) hybridizes BPMN and threat modeling, further subjected to model transformations into RDF graphs. A demonstrator is implemented on the ADOxx metamodeling platform and its interoperability/adapters for integration with triplestores and LLMs—this makes possible both SPARQL and natural language queries over the prescriptive diagrammatic designs, as analysis approaches. The proposed method can be used to describe contextualized cybersecurity threats during the design phase of a system or during IT system auditing. The integration bridge with knowledge graphs and large language models enhances the system analysis capabilities by streamlining heterogeneous knowledge flows comprising diagrammatic representation, RDF graphs and generative AI, towards an emerging flavor of model-driven engineering. The method was developed iteratively according to the Design Science framework, with its development activities specialized for the DSML-focused Agile Modeling Method Engineering framework.