Post: Predict Batch

Endpoint: /pyspark_prediction/predict/batch/

Executes batch prediction in parallel using PySpark for large input datasets. This enables fast inference at scale by distributing processing across Spark workers.

Input Parameters:

• usecaseId (string, required): Unique identifier for the trained use case/model.
• usecaseVersion (integer, optional): Version of the use case (default: 1).
• Request Body (multipart/form-data):

  • • • file (binary, required): Input CSV or structured file containing feature records for batch prediction.

        Output:

        Returns a JSON response with a reference to the prediction output file or the results, depending on configuration.