I Built a Production-Grade Async Job Queue from Scratch — Here's Everything That Actually Happened

Built a production-grade async job queue from scratch with Python, FastAPI, Redis Streams, and PostgreSQL, instead of relying on Celery.

The system adds backpressure, priority scheduling, crash recovery, and a Reaper service to handle timed-out and zombie jobs.

The project was developed in eight validated stages and finished with 47 passing tests, 85% coverage, and a mostly complete services layer.

It offers a practical blueprint for designing reliable backend queues and preparing for systems design interviews.