Can Virtual Agents Care? Designing an Empathetic and Personalized LLM-Driven Conversational Agent

TL;DR

This paper presents a virtual agent framework using retrieval-augmented architecture to deliver empathetic, personalized, and reliable mental health support, outperforming baseline models in cross-cultural evaluations.

Contribution

Introduces a novel virtual agent framework combining retrieval, structured memory, and multimodal interaction for empathetic mental health support.

Findings

Improved retrieval and response quality for smaller models.

Outperforms LLM-only baselines in coherence, accuracy, and empathy.

Participants prefer the proposed system over baseline models.

Abstract

Mental health challenges are rising globally, while traditional support services face limited availability and high costs. Large language models offer potential for conversational support, but often lack personalization, empathy, and factual grounding. A virtual agent framework is introduced to provide empathetic, personalized, and reliable wellbeing support through retrieval-augmented architecture, structured memory, and multimodal interaction. Objective benchmarks demonstrate improved retrieval and response quality, particularly for smaller models. A cross-cultural study with university students from Vietnam and Australia shows the system outperforms LLM-only baselines in coherence, perceived accuracy, and empathy, with most participants clearly preferring the proposed approach.