Describe the last script you wrote that ran on a schedule (e.g., cron, Airflow) rather than on your laptop. What did it do from start to finish?

Situation: In my last role, I built a nightly ETL script that extracted user activity logs from an S3 bucket, transformed them into a denormalized format, and loaded them into a Snowflake table. Task: The script had to run automatically every midnight without manual intervention. Action: I wrote the script in Python, using the boto3 library to list and download the latest log files, pandas for transformation, and the snowflake-connector-python to load the data. I packaged the script into a Docker container and scheduled it with Airflow’s DAG, setting a cron expression for 00:00 UTC. I added error handling to retry failed downloads, logged progress to CloudWatch, and sent a Slack notification on success or failure. Result: The scheduled job ran reliably every night, processed 10 GB of logs in under 15 minutes, and the data was available in Snowflake for downstream analytics. The automation reduced manual effort by 80 % and improved data freshness for the BI team.