I’d suggest automating the process using a Cloud Run function that is triggered whenever a new CSV file is dropped into the GCS bucket. You can use a Python or Node.js script to read the files and transform them. Afterwards, load the data using the appropriate driver or connector. You can read into this article regarding connecting to AlloyDB for PostgreSQL.

If you want to focus more on scheduling the process, consider using Cloud Scheduler. Here’s the documentation for scheduling a cloud run function.