AI-driven solutions are transforming telecommunications, but their lack of transparency makes them hard to trust. While XAI techniques help explain AI decisions, their outputs are often too technical for telecom professionals to interpret. This seminar explores an ongoing research effort to use Large Language Models (LLMs) to simplify AI explanations. By converting complex AI insights into clear, structured narratives, LLMs make decision-making more understandable and actionable. Our approach integrates network data, decision flows, and key performance indicators (KPIs) to improve trust and usability in AI-driven telecom systems. Early results suggest that LLM-driven explanations enhance human understanding, supporting better AI adoption in telecom.

About Andrea Fresa

Andrea Fresa holds a Master’s and Bachelor’s degree in Computer Science Engineering from Università degli Studi di Napoli Federico II and is currently pursuing a PhD at IMDEA Networks Institute and Universidad Carlos III de Madrid. His research focuses on Edge Computing and AI. As a research intern at Ericsson Research, he worked on robotic service placement in the Edge-Cloud continuum and IoT intercommunication. Currently, he is exploring the use of Large Language Models (LLMs) to enhance the explainability of AI solutions in telecommunications.

This event will be conducted in English