Carmelo Fabio Longo currently serves as Chief Information Security Officer (CISO) at the University Hospital Policlinico of Catania (Italy), where he is responsible for cybersecurity strategy, information security governance, risk management, and the protection of critical healthcare IT infrastructures. After graduating in Computer Science in 2000, he began his professional career in healthcare IT engineering, gaining extensive experience in the design, management, and security of complex information systems. Since 2017, he has also been involved in applied research on Artificial Intelligence, robotic and multi-agent systems, with a particular focus on natural language processing and cognitive architectures.

In 2022, he obtained his Ph.D. in Computer Science from the University of Catania. He subsequently worked as a researcher at the Institute of Cognitive Sciences and Technologies of the Italian National Research Council (CNR-ISTC), conducting research on artificial cognitive systems, semantic knowledge representation, and intelligent agents.

His main research interests include Artificial Intelligence, Cognitive Architectures, Natural Language Processing, Multi-Agent Systems and Semantic Web. 

Scientific curricula: https://www.dropbox.com/scl/fi/vltrl4b2pg7cqm7pxfuq7/Curricula.pdf?rlkey=btfj7vsoyo4phnudmdyc2ugg1&dl=0

The article “A privacy-preserving chatbot framework leveraging open-source large language models for patient simulation in psychotherapy training”, published in Health Psychology Research in September 2026, introduces Psy-Mimic, an innovative platform leveraging open-source Large Language Models (LLMs) to simulate virtual patients for psychotherapy training.

The framework aims to provide psychotherapy trainees with a safe, controlled, and realistic environment in which to practice and refine their clinical skills through interactive conversations with AI-generated patients. A key distinguishing feature is its strong emphasis on privacy preservation, enabling local deployment without relying on external cloud services.

Developed through an interdisciplinary collaboration combining artificial intelligence, computer science, and clinical psychology, Psy-Mimic represents a promising step toward scientifically validated AI-driven psychotherapy training tools, laying the groundwork for future advances in computational patient simulation, behavioral realism, and multimodal interaction.

Paper: https://healthpr.org/journal/HPR/articles/online_first/9211

Joyful moments from the presentation of Psy-Mimic by the validation team led by Prof. Pasquale Caponnetto. The Congress provided a valuable opportunity for the team to engage directly with mental health professionals, fostering insightful discussions on the platform’s effectiveness and potential applications in psychotherapy training. The event was a great success, enabling the team to gather valuable feedback, exchange perspectives, and identify promising directions for future developments and improvements of the Psy-Mimic platform.

Pictured from left to right (upper picture): Dr. Giulia Schilirò, Dr. Noemi Vitale, and Dr. Alessia Sangiorgio, all members of the Psy-Mimic team, whose contributions are instrumental to the project's validation and ongoing development.

Among LLMs meta-cognitive techniques based on chain-of-thoughs, here we investigated whether metaknowledge can be injected by fine-tuning in an auto-regressive fashion-way based on past inferences. 

https://doi.org/10.1016/j.cogsys.2025.101352 

The conference was featured by Jurgen Schmidhuber keynote, who can be considered the father of the pre-transformers state-of-the-art architecture known as Long Short Term Memory (LSTM)