Will ray data support more preprocessors？

Ray Libraries (Data, Train, Tune, Serve)
1

I’m new to Ray, compared to spark ML lib, I found the kinds of preprocessors are much fewer… that’s make it hard to do distributed data processing in the same Ray cluster when facing feature-engineering works