arulsebastin71/smartqueue

SmartQueue is an AI-powered queue management system built as an IEEE final-year project. Its purpose is to reduce waiting time for customers by letting them book time slots in advance, track their position in a live queue, and receive a QR-code token that represents their turn. The system is divided into three parts: a machine learning service, a backend API, and a React-based web frontend. The machine learning component is a Python service that predicts how long a customer will wait for any given time slot. It uses a gradient boosting model trained on 8,000 samples and takes into account factors like the hour of the slot, the day of the week, current queue length, average service time, whether it is a peak hour, recent booking volume for that slot, and the proportion of elderly visitors. The model reports an R-squared value of 0.9985 and a mean absolute error of about 2.3 minutes, meaning its wait-time predictions are typically within about two minutes of the actual outcome. The backend handles user registration, bookings, and the queue itself. It runs on Node.js with Express and stores data in MongoDB. An admin role can be assigned manually through the database. The backend also implements a priority scoring system called Tribonacci priority that dynamically adjusts each booking's position based on how long the person has already been waiting. When a business has a long queue, the system can suggest nearby alternatives in the same category with shorter queues. The frontend is built with React and includes a QR code scanner that reads booking tokens via a phone camera. Admins have access to a dashboard with a heat map of peak hours built from 30 days of booking history, which helps visualize busy periods at a glance. The README focuses on setup instructions and technical specifications rather than a general description, so this is primarily a demonstration project aimed at academic evaluation rather than production deployment.