[alexott]

there is Pandas on Spark, included into Spark itself (originally Koalas) - the switch to it is very easy, and you get parallelization.