ECHO: A Locally-Deployable Agentic Health Assistant with Temporal Memory, Safety Guardrails, and Speech Assessment
ECHO is a locally deployable chronic-care health assistant that combines a ReAct/LangGraph clinical-tool agent with persistent temporal memory, a hybrid safety filter for crisis/jailbreak/clinical-boundary detection, and a multimodal speech module for estimating emotion, depression, and pain. What’s interesting is the end-to-end packaging of memory, guardrails, and speech assessment into a privacy-preserving consumer-hardware app, but the evidence is still fairly prototype-like: the tool benchmark is small, the safety model is validated on a Turkish dataset, speech performance is moderate, and the abstract’s use of GPT-5 Mini raises questions about how fully local the strongest configuration really is.
Abstract
This paper presents ECHO (Enhanced Care \& Health Observer), a locally-deployable conversational health assistant for long-term chronic care management. ECHO integrates three complementary software modules developed under shared supervision as a unified system. The core module is an agentic chatbot built on a ReAct loop orchestrated via LangGraph, equipped with 17 clinical tools and a temporal knowledge graph for persistent cross-session memory; it achieves a 94.9\% tool-execution pass rate across a 59-scenario benchmark with GPT-5 Mini. A two-stage hybrid safety layer intercepts all incoming queries: a rule-based layer handles explicit crisis signals and jailbreak attempts in under 1ms, while a signed graph neural network (GNN) with APPNP-style propagation classifies boundary cases by clinical intent, achieving 88.8\% accuracy and 90.6\% unsafe recall on a 2,537-query annotated Turkish health dataset while outperforming zero-shot LLM baselines including Llama 3.3 70B. A multimodal speech assessment module combining Whisper acoustic encoding and BERT text encoding with cross-attention fusion estimates emotion, depression, and pain, reaching a mean macro F1 of 0.652. The full system is implemented as a web application that can run entirely on consumer hardware, with no patient data transmitted to external services, supporting compliance with GDPR and KVKK.