IssueTracker

What you would like to accomplish:

I would like to be able to use BigQuery views as input data for a Vertex AI Batch Prediction job.

How this might work:

I don't know much about the internals of Batch Prediction Jobs, so I'm not sure.

If applicable, reasons why alternative solutions are not sufficient:

There are cases where it's inconvenient to create a table just to serve as an input to a prediction job - you have to take care of cleanup later on yourself. Additionally, it doesn't always make sense to keep this table indefinitely because the same data will be available in the prediction job output either way.

Other information (workarounds you have tried, documentation consulted, etc):

My workaround is to write all my data into a table, run a prediction job and then delete the table.