Ask HN: Who is using spark/parquet nested schema?

After having seen Spark + Parquet In Depth (https://www.youtube.com/watch?v=_0Wpwj_gvzg), I have found very inspiring the nested schema feature. I have a lot of use case in mind, where this could make manipulating data more readable.

Now coming from "regular data" world (Oracle), I am still a bit worried about using it. So my question: are there people here using such nested schema, with reasonable volume of data in production like system.

What are your experience with spark and parquet nested schema, is there any specific issue to be aware of ? Any regrets to have tried to use it ? Or on the contrary a good bet ?

Edit: add Ask HN in title.