Feature Description

Will it ever be possible to run ingestion in parallell with the parsing. For ex. by llamaparse emitting an event for every page which a worker can handle

Reason

No response

Value of Feature

Current recommended approach is to poll until the full document processing is done, which is inefficient.