How to Use Melisyrch to Ingestion Documents

I'm curious how you're handling the ingestion process of Melisyrch. They have this, I'm seeing in the documentation here, they have this internal queuing system for ingesting documents with tasks. What's been your experience using that? It's almost like the millisearch team saw what Andrew came was doing and loved it then just re-implemented it with their own version of a much more simple elastic search.