Multi-agent chatbot for early Chronic Kidney Disease using FLOWISE no-code platform, democratising health software development.

Abstract

Chronic Kidney Disease (CKD) is a global health challenge often diagnosed late, when opportunities for slowing progression are limited. Early lifestyle intervention and patient self-management can substantially improve outcomes. This study presents a multi-agent chatbot system designed to support early CKD intervention, developed using the no-code Flowise AI platform. The approach demonstrates how no-code tools can democratise software development, enabling the rapid creation of intelligent health applications without extensive programming expertise. System requirements were derived from a review of CKD self-management apps and refined through MoSCoW prioritisation, covering dietary guidance, symptom monitoring, exercise, medication adherence, and education. While an initial Figma prototype was used to visualise user requirements, the implementation focused on a multi-agent chatbot comprising a Supervisor, Diet Coach, and Symptom Manager, integrated with KDIGO clinical guidelines. Evaluation with 12 healthcare professionals produced positive feedback, with mean scores above 4.2/5 for usefulness, usability, and perceived accuracy. The study highlights the potential of combining multi-agent architectures with no-code development to deliver accessible, personalised digital health tools. Future work will expand the agent set, integrate patient data, and evaluate the system in clinical contexts.