PersonaPlugin: A Multi-Source Persona Framework for LLM Personalization in Telecommunications

Personalizing large language model (LLM) responses in enterprise environments faces unique challenges: heterogeneous data sources, strict on-premise constraints, and unreliable LLM outputs. We present PersonaPlugin, a framework deployed in a live telecom environment that transforms real-world behavioral signals—call summaries, app usage, and location patterns—into structured personas for LLM personalization. Our system processes 1M+ behavioral records daily, entirely on-premise. The framework comprises three specialized engines, with the call feature extraction engine achieving $>$99% parsing success through cascaded fallback mechanisms. We conduct a comprehensive two-track evaluation: (1) general personalization across 800 scenarios, and (2) call-context personalization across 400 scenarios. Our LLM-as-a-Judge evaluation, validated against human judgments ($\kappa=0.71$), demonstrates that persona integration significantly improves personalization: +24.0% in P-Score and +26.7% in Adherence over baseline. Integrating call context with personas yields +11% additional improvement, validating the synergistic value of multi-source persona construction.