Queue Cure '26 — Live Clinic Queue Manager

76% of India's 1.5 million neighbourhood clinics run on paper token slips and verbal announcements. Patients endure 2-3 hour waits with zero visibility into their actual wait time. Receptionists manage the entire queue from memory with no dashboard or data. This affects millions of patients daily across the country. The core pain points are: patients have no idea when they will be called, receptionists have no system to manage flow efficiently, and there is no data collected on wait times or consultation duration. Process I started by mapping the core user flows — receptionist adding patients and calling tokens, patients checking their wait status. I chose Socket.IO over HTTP polling after researching real-time communication options, because WebSocket delivers sub-second updates without wasting server resources. I built the backend first with Express and MongoDB, tested all API routes using Thunder Client, then built the React frontend. I faced a concurrency challenge where two patients could get the same token number — I solved this using atomic MongoDB operations. I also struggled with Socket.IO CORS issues on deployment, which I fixed by configuring the server to allow all origins explicitly. Results Live queue updates are delivered in under 1 second via WebSocket across both screens with zero page refresh. A receptionist can add a patient and assign a token in under 10 seconds. The estimated wait time recalculates dynamically every time a token is called or the average consultation time is updated — never hardcoded. The app is fully deployed and live at queue-cure-fawn.vercel.app. Both screens tested simultaneously on separate devices confirmed real-time sync working correctly. Reflection I would add a doctor-facing screen from the start that lets the doctor mark a consultation as complete, triggering automatic token advancement instead of manual receptionist action. I would also implement SMS or WhatsApp notifications via Twilio so patients can wait outside and get alerted when their token is two slots away. On the technical side, I would use Redis as a pub-sub layer instead of in-memory Socket.IO for multi-server scalability, and add a proper authentication system instead of a hardcoded clinic password.