Thank you for this presentation. I am working on giving users a way to import data and this really helped clear a path forward for this idea. The discussion around cleaning data was very helpful as that is what I struggle with the most.
I am thinking that maybe I should even do this data ingestion on a separate Dgraph instance from my main running instance and then link the data and clean the data and then merge the imported data from the temp Dgraph instance to my primary running Dgraph instance. Given the ability to start up a new backend on Cloud and kill it after the process completes with the API, is this the method you would recommend as well?