Course Wiki
Deployment Automation
From Colab notebooks to scheduled Python: files, SQLite storage, cron, and Airflow.
Related Pages
Module 5 closes the loop from analysis to operation. Notebooks become Python files; prediction outputs get persistent storage (files or a simple SQLite database, with an SQL introduction); and the whole prediction pipeline runs on a schedule — cron for a series of scripts, with Apache Airflow introduced as the workflow-level alternative.
The automation target is the full cycle: download fresh data, compute indicators, generate predictions, execute or log the trades the strategy calls for — systematically, on schedule, without a human opening a notebook. An optional track adds email notifications with predictions and profit/loss updates.