The research report “A Retrieval Augmented Generation-based technique to guide firewall configuration with Large Language Models” has been accepted for publication

The research report “A Retrieval Augmented Generation-based technique to guide firewall configuration with Large Language Models” has been accepted for publication in Springer’s prestigious “Lecture Notes in Artificial Intelligence” series, in the volume that will be available from December 2025.

The research, carried out by colleagues Roberto Lorusso, Antonio Maci and Antonio Coscia, of the CyberLab BV TECH as part of the activities envisaged by the Program Contract “Cybersecurity and SOC Product Suite“, it presents in detail how the architecture of a conversational assistant has been designed and validated which combines a compact Large Language Model (LLM) with specialized Retrieval-Augmented Generation (RAG) techniques to support IT security equipment configuration activities, and capable to operate effectively in disconnected mode.