SeatFinder

Overview

SeatFinder is a course enrollment tracking service designed to help university students secure spots in full classes. We actively monitor course and waitlist availability and notify our students the moment a seat becomes available.

We monitor seat availability across major universities in Canada: McGill, U of T, Concordia, McMaster, Dalhousie, UdeM, uOttawa, and Western. Since launching in 2025, we’ve sent over 31,000 notifications across 2,300 tracking requests.

Architecture

The core challenge is monitoring thousands of courses simultaneously with minimal latency. We use Celery, a distributed task queue, to parallelize course checks across multiple workers. Each worker polls a batch of courses, compares availability against its cached state, and triggers notifications on changes.

This architecture scales horizontally: adding workers increases throughput linearly. Tasks are distributed via a message broker (Redis), ensuring no single point of failure and enabling sub-second response times even under heavy load.

Stack: Flask, Celery, Redis, MongoDB