Seems like it should be. That's pretty much the status quo with how Azure Synapse works. You put your Parquet file out on storage, and query it (via Spark or Pandas).