Mohamed Jaasir B

UPriTer Med: AI-Powered Real-Time Queue & Arrival Planning System for Multi-Speciality Hospitals

Multi-speciality hospitals serve hundreds of patients daily, yet many still spend 1–3 hours waiting despite having appointments. Patients lack visibility into queue status and doctor delays, receptionists manually manage schedules and repeated queries, doctors face uneven workloads, and administrators lack operational insights. Existing platforms solve booking but not post-booking patient flow. Upriter addresses this gap with AI-powered scheduling, live queue tracking dynamic wait prediction, and personalized arrival recommendations to reduce unnecessary waiting and improve hospital efficiency Process I began by studying existing hospital workflows and realized that appointment booking doesn't eliminate waiting. My first idea was a basic token system, but it felt too similar to existing solutions. I reframed the problem from "managing queues" to "reducing unnecessary waiting." I mapped four stakeholders—patients, doctors, receptionists, and administrators—and designed separate dashboards for each. I initially considered fixed wait times, but rejected them because consultation durations vary. Instead, I used actual consultation data for predictions. Features like payments and telemedicine were intentionally excluded to keep the solution focused. The final addition was the AI Arrival Planner, which tells patients when they should leave home based on real-time queue status. Results The final prototype successfully connected four stakeholders through a unified workflow with real-time synchronization and no page refreshes. Dynamic wait-time prediction replaced fixed estimates by using actual consultation durations. The AI Arrival Planner transformed queue tracking into patient journey planning by recommending when patients should leave home. If given more time, I would validate the system with real hospital staff and patients, collect usability metrics, and further improve prediction accuracy using larger datasets. Reflection Given more time, I would validate the product with real patients, doctors, and reception staff to better understand their workflows and pain points. I would improve the AI prediction model using larger consultation datasets and add notification channels such as WhatsApp reminders. I would also introduce support for multi-doctor dependencies, lab visits, and follow-up journeys. During development, I intentionally prioritized solving the waiting problem first instead of expanding into payments, telemedicine, or insurance workflows.