Transform your raw input data using a saved query before requesting predictions and then use ML.EVALUATE, after defining preprocessing steps with BigQuery‘s TRANSFORM clause during model creation.

Define your preprocessing logic using a BigQuery view for model training data. At prediction time, use BigQuery‘s ML.EVALUATE clause without any transformations on the raw input data.

Preprocess all data using Dataflow. At prediction time, use BigQuery‘s ML.EVALUATE clause without any further transformations on the input data.

Use BigQuery‘s TRANSFORM clause to define preprocessing steps when creating your model. Then, at prediction time, use BigQuery‘s ML.EVALUATE clause without specifying any transformations on the raw input data.