An Intelligent Cloud-Deployed Chatbot for Automating Hospital Front-Desk Enquiries with Intent-Aware Dialogue Management

Abstract: Front-desk operations in healthcare institutions absorb a continuous influx of repetitive questions about scheduling, departments, timings, and procedures, the manual servicing of which burdens reception staff and prolongs the time patients spend waiting for answers. Software agents that comprehend everyday language can shoulder much of this load, yet a considerable number of fielded healthcare bots still depend on rigid scripts that falter when wording shifts, and many run on fixed servers that cannot accommodate spikes in demand. This work describes an intelligent, cloud-deployed chatbot that reads unconstrained user messages, infers the user's goal together with the relevant details, and either answers directly or launches the corresponding task such as scheduling a consultation. The design unifies a learned intent recognizer, an entity extractor, and a context-tracking dialogue controller, and deliberately routes uncertain exchanges to reception staff to safeguard quality. A React and Node.js conversational front end is paired with a Python language-processing core and hosted on a public cloud that supplies on-demand capacity and unified monitoring. On a purpose-built collection of hospital-domain messages, the intent recognizer achieved 94% accuracy and entity extraction reached an F1 of 0.91, while the hosted service held median replies under one second during concurrent usage. Against a scripted baseline, the proposed agent showed clear improvements in resilience to paraphrase and in user satisfaction. The contributions comprise a domain-specific understanding-and-dialogue pipeline, an uncertainty-triggered staff handoff, and a measured cloud deployment balancing speed, uptime, and ease of operation.

Keywords: Chatbot, Natural Language Understanding, Intent Recognition, Dialogue Management, Healthcare Automation, Cloud Deployment, Entity Extraction, Patient Services