Knowledge-Guided Agentic Framework Mitigates Medical Query Ambiguity

An agentic framework uses task-specific knowledge graphs to resolve patient-context ambiguity by identifying missing variables like allergies or medications. It generates targeted follow-up questions to distinguish between plausible hypotheses before the downstream LLM generates a final medical response.