QueueCure AI — Real-Time Clinic Queue Management System

76% of India's 1.5 million clinics run on paper token slips. Patients wait 2–3 hours with zero visibility into when they'll be called. Receptionists track everything manually. Doctors have no operational dashboard. This causes frustration, walkaways, and inefficient clinic flow — all solvable with real-time software. Process Started by mapping the core pain: patients don't know their wait time. Designed the queue engine first as pure functions (no Express dependency) so logic could be tested in isolation. Built Socket.IO sync before UI so live updates worked from day one. Iterated on the Clinic Pulse formula — tried averages first, switched to weighted scoring. The emergency auto-priority behavior was a late addition after realizing a fixed position insert was more reliable than sorting. Built with AI assistance using Claude (Anthropic) for code architecture, component structure, and iterative debugging — treating it as a pair programmer throughout the build. Results Delivered a fully working full-stack system in 4 days. Both screens sync in under 1 second via WebSockets. Emergency patients auto-prioritize to position 1. 13/13 unit tests passing. Analytics dashboard shows peak hour, pulse trend, and consultation type breakdown. Deployed backend on Railway, frontend on Vercel. This project was built with Claude (Anthropic) as an AI pair programmer. Prompting, reviewing, and directing the AI was itself a core skill demonstrated — knowing what to ask, when to push back, and how to validate AI-generated code with unit tests. Reflection Add SMS notifications so absent patients get a reminder before their token is called. Implement actual MongoDB time-series logging for the Clinic Pulse trend instead of approximating it from current state. Would also add a doctor-side view showing their personal consultation history and average session time.