Kafka or similar is already used by the existing solutions mentioned in the post.

I think what this adds is a better way of dump processing without using database specific features.