José Vázquez Naya presents AI advances on traffic management and forensic analysis at the Universidad Alfonso X el Sabio
A Coruña, March 16th, 2026
The CITIC researcher at the UDC, José Vázquez Naya, participated this past February as a guest lecturer in the Conferences on AI and Cybersecurity Research, held at the Universidad Alfonso X el Sabio. The event was a part of the PECIEE project, a collaborative initiative between the Universidad de Oviedo, the Universidad Alfonso X and INCIBE – the National Cybersecurity Institute.
During his presentation, titled “AI Applications in Cybersecurity: Network Traffic Management and Identification of Forensic Artifacts,” Vázquez Naya presented the latest advances made by his team at the CITIC, demonstrating how Artificial Intelligence is transforming computer security and digital forensics.
One of the central themes of his presentation was the evolution of asset discovery techniques, with a particular focus on operating system fingerprinting. The researcher explained how his team has transitioned from using classic machine learning algorithms, embodied in the open-source tool fingerAI, to employing tabular architectures based on Transformers—advances that have already been validated and accepted for publication in the academic journal Cybersecurity.
As a major future endeavour, he presented the NetHermes project, aimed at revolutionizing Security Operations Centres (SOCs) through the development of the first multimodal traffic and language model. This model allows for the native processing of network packets along with natural language instructions, overcoming the current “textual bottleneck” and facilitating much more precise analytical interactions. The state-of-the-art research supporting this innovation was recently published in the journal Computer Networks.
n the field of computer forensics, Vázquez Naya showcased an AI module developed for the Autopsy platform, which integrates artificial vision models such as CLIP and YOLO. This solution enables semantic searches, visual similarity analysis, and automatic tagging of large volumes of seized images, increasing the efficiency of the analysis and ensuring complete legal traceability of the evidence.
The presentation also highlighted the group’s strong international focus, with active research and collaborations with leading European laboratories, including work on conformal prediction with KCLIP at King’s College London and the development of advanced authentication systems in collaboration with the SSL at the Universidad del Pireo.
