CoRDial

This project successfully developed a multi-agent conversational framework designed to enhance human-robot interactions in elder care settings. Through seven co-design workshops with older adults and care workers in urban, rural, and respite care environments, we gathered valuable insights into user needs, leading to key refinements in our approach. The main findings from the workshop were around language/vocabulary, personalisation, data protection and safety.

Our framework, built using OpenAI services and Langchain, features a routing manager based on prototypical networks that dynamically assigns user queries to the most relevant task agent. These advancements contribute to making conversational AI more adaptive, transparent, and user-centered, enhancing its effectiveness in elder care settings.

CoRDial offers a flexible and modular system for integrating new agents as tasks evolve. The Dialogue Manager oversees interactions, routing queries to specialized agents using the ProtoNet Router. This ensures effective management of user requests while maintaining context and handling ethical considerations. The Ethical Reviewer ensures compliance with healthcare guidelines. Task agents, l ike Off-Topic Question Answering and Medication Reminders, operate with unique personas and access relevant APIs.

Each agent has dedicated episodic memory for retaining conversation histories and providing context. The ProtoNet Router uses few-shot learning to classify user responses, ensuring efficient agent routing with improved interpretability. The system also integrates continuous learning, adapting based on user feedback stored in Firebase Cloud Firestore to enhance routing accuracy and overall conversational quality over time.

The CoRDial dashboard offers users a range of functionalities, including access to an inbuilt chatbot for non-simulator tasks and the ability to manage patient personas.