Bridging Agility and Automation: Enhancing Model-Driven Engineering with LLMs in Scrum

Leila Samimi; Shekoufeh Kolahdouz Rahimi

Bridging Agility and Automation: Enhancing Model-Driven Engineering with LLMs in Scrum

Leila Samimi
, Shekoufeh Kolahdouz Rahimi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

52 Downloads (Pure)

Abstract

By concentrating on high-level models that simplify design and automate coding, Model-Driven Engineering (MDE) assists developers in tackling challenging software projects. However, in projects that are dynamic and fast-paced, its rigid procedures may seem cumbersome. Scrum and other agile methods provide the necessary flexibility, but integrating them with MDE can be challenging because modeling frequently requires labor- intensive manual work. To make MDE more agile within Scrum, we suggest utilizing Large Language Models (LLMs), which allows these intelligent tools to take care of tasks like building models or checking their accuracy. Through a case study on a ticket reservation system, we show how LLMs can speed up modeling and make sprints more adaptable. While hurdles like ensuring LLMs get domain details right or syncing them with tools like Papyrus remain, our approach paves the way for combining precision of MDE with responsiveness of Agile, with plans to test it on larger, more complex systems.

Original language	English
Title of host publication	Agile Model-driven Engineering Workshop
Publisher	CEUR Workshop Proceedings
Publication status	Published - 2025

Access to Document

Samimi_STAF_2025_paper_91

Cite this

@inproceedings{2aa379ae589a44bc9df9f166d7e8d1f3,

title = "Bridging Agility and Automation: Enhancing Model-Driven Engineering with LLMs in Scrum",

abstract = "By concentrating on high-level models that simplify design and automate coding, Model-Driven Engineering (MDE) assists developers in tackling challenging software projects. However, in projects that are dynamic and fast-paced, its rigid procedures may seem cumbersome. Scrum and other agile methods provide the necessary flexibility, but integrating them with MDE can be challenging because modeling frequently requires labor- intensive manual work. To make MDE more agile within Scrum, we suggest utilizing Large Language Models (LLMs), which allows these intelligent tools to take care of tasks like building models or checking their accuracy. Through a case study on a ticket reservation system, we show how LLMs can speed up modeling and make sprints more adaptable. While hurdles like ensuring LLMs get domain details right or syncing them with tools like Papyrus remain, our approach paves the way for combining precision of MDE with responsiveness of Agile, with plans to test it on larger, more complex systems.",

author = "Leila Samimi and \{Kolahdouz Rahimi\}, Shekoufeh",

year = "2025",

language = "English",

booktitle = "Agile Model-driven Engineering Workshop",

publisher = "CEUR Workshop Proceedings",

}

TY - GEN

T1 - Bridging Agility and Automation: Enhancing Model-Driven Engineering with LLMs in Scrum

AU - Samimi, Leila

AU - Kolahdouz Rahimi, Shekoufeh

PY - 2025

Y1 - 2025

N2 - By concentrating on high-level models that simplify design and automate coding, Model-Driven Engineering (MDE) assists developers in tackling challenging software projects. However, in projects that are dynamic and fast-paced, its rigid procedures may seem cumbersome. Scrum and other agile methods provide the necessary flexibility, but integrating them with MDE can be challenging because modeling frequently requires labor- intensive manual work. To make MDE more agile within Scrum, we suggest utilizing Large Language Models (LLMs), which allows these intelligent tools to take care of tasks like building models or checking their accuracy. Through a case study on a ticket reservation system, we show how LLMs can speed up modeling and make sprints more adaptable. While hurdles like ensuring LLMs get domain details right or syncing them with tools like Papyrus remain, our approach paves the way for combining precision of MDE with responsiveness of Agile, with plans to test it on larger, more complex systems.

AB - By concentrating on high-level models that simplify design and automate coding, Model-Driven Engineering (MDE) assists developers in tackling challenging software projects. However, in projects that are dynamic and fast-paced, its rigid procedures may seem cumbersome. Scrum and other agile methods provide the necessary flexibility, but integrating them with MDE can be challenging because modeling frequently requires labor- intensive manual work. To make MDE more agile within Scrum, we suggest utilizing Large Language Models (LLMs), which allows these intelligent tools to take care of tasks like building models or checking their accuracy. Through a case study on a ticket reservation system, we show how LLMs can speed up modeling and make sprints more adaptable. While hurdles like ensuring LLMs get domain details right or syncing them with tools like Papyrus remain, our approach paves the way for combining precision of MDE with responsiveness of Agile, with plans to test it on larger, more complex systems.

M3 - Conference contribution

BT - Agile Model-driven Engineering Workshop

PB - CEUR Workshop Proceedings

ER -