So at this stage you would need to pipe the Hive data into Grakn.
However, we are at the moment trying to identify other tools to more strongly couple with. We already are using Apache Spark for our analytics platform but nothing we have "out of the box" can automigrate your existing Hive data.
If you were able to convert your data into CSV or JSOn then you could migrate using one of our existing tools.
If you can't do that then you would need to create a custom pipeline as you mentioned. For this we have a batch loading client which may be able to help.
Hope that's a start at least.