A serious data start-up structured around a Julia data manipulation framework for larger-than-RAM data

Yes, fast user-defined functions is one of the advantages that a Julia solution would have over Polars or Spark.