QueueCure — Real-Time Hospital Queue Management System

Hospital outpatient departments in India run on paper tokens and verbal announcements — patients sit anxiously for hours with no visibility into how long they'll actually wait. There's no way to know if you're 2 patients away or 20. Receptionists juggle registers, sticky notes, and shouting names across crowded waiting rooms. The core pain points: • Patients can't leave to grab water or use restrooms without risking missing their turn • Receptionists have no single dashboard — no live count, no history, no skip tracking • Wait time estimates are guesswork ("around 45 minutes, sir") I started by mapping the two-sided user flow: receptionists who manage state, and patients who are passive consumers of information. These needed fundamentally different UIs — a command center vs. a calm display board. Architecture decision: I chose WebSocket (Socket.IO) over polling because queue state changes are event-driven and latency matters — a patient shouldn't see stale "Token 12" while the screen outside says "Token 14". Every mutation (add, call next, skip, complete) emits a `queueUpdated` event that instantly refreshes all connected clients. For wait time, I built a dual-average system: a configurable fallback (default 8 min) bootstraps the first patients of the day, then a dynamic average recalculates from actual completed consultation durations logged in MongoDB. The live system handles the full outpatient loop: registration → token assignment → live queue display → call next → complete → dynamic time update, all propagated in under 100ms via WebSocket. Key outcomes: • Dynamic wait estimates replace static guesswork — accuracy improves throughout the day as real durations are logged • Receptionists have a single dashboard: add patient, call next, skip, complete, and view history — all in one view •patients can self-check their token position without approaching the counter • Skipped patients are tracked separately according to estimated time. Add real-time SMS/WhatsApp alerts for patients 2 tokens away. Split queue per department for multi-doctor clinics. Add auth for receptionist dashboard (currently open access). Persist token sequence across day resets using a daily counter in MongoDB.