User complaints that upload of large batches is too slow

During training in Kenya, users complained that the upload of large user batches takes too long.

Can we understand how to make this faster?
Should we increase parallelization?
Should we have true independent parallelization instead of waiting for the batch to complete before starting another?
Could we use the bulkDocs endpoint to upload users in a single request?

medic / cht-user-management