QueueCure - Real Time Smart Clinic Queue Management System

Waiting in clinic reception rooms is frustrating, stressful, and leads to severe overcrowding. Patients have zero visibility into actual wait times, while receptionists are overwhelmed by manual queue tracking, calling tokens, and handling priority cases. This lack of transparency causes high patient churn and operational chaos. We needed a system that replaces paper tokens with dynamic digital wait times, keeping patients updated on their mobile devices so they can wait comfortably elsewhere while giving clinics real-time management and analytics. Process I began by mapping out the patient journey and clinic workflow. I realized that static token numbers cause anxiety; patients need an active 'SmartReturn' guidance system. I designed a Socket.io-backed real-time architecture to instantly sync state changes. I built the Frontend with Vite React and CSS variables for smooth theme transitions, and the backend using Node/Express/MongoDB. I prototyped two versions: first with static estimated wait times, and later dynamic estimations using moving averages of consultation durations. During testing, we encountered connection recovery issues when mobile screens went dark. I solved this by implementing state recovery protocols in Socket.io to resume session data seamlessly upon reconnecting. Results I successfully built a responsive web application that synchronizes state changes across the Clinic Console, Patient Tracker, and Public Display in under 100ms. The SmartReturn algorithm successfully adapts to doctor pace by updating estimated wait times dynamically. In simulated user tests, real-time browser alerts successfully notified patients when they were 2 tokens ahead, eliminating waiting room overcrowding. The owner analytics dashboard successfully tracked hourly patient flow and average consultation times, allowing clinics to optimize scheduling. Reflection In retrospect, I would integrate native SMS/WhatsApp API notifications directly instead of relying solely on browser push alerts, as web notifications can be blocked by certain mobile browsers. Additionally, I would implement machine learning models (like regression trees) to predict wait times using historic daily trends (e.g., peak hours, visit types) rather than relying solely on a simple moving average of the current session, which would make wait predictions even more accurate during morning rushes.