I would like to know the tools which are available for filtering the read level data in RNA-seq after the alignment based on the duplicates, read metrics etc.
featureCounts have options such as ignoreDup or countMultiMappingReads which allows you to perform some filtering (I am assuming you have BAM files. Please provide more details otherwise). However, generally you do not want to remove duplicates in RNA-seq. There are several discussions on this, and here is an older thread that has some discussion which could be useful to you.
And just to ensure I covered this: some of this filtering should be done prior to alignment. You mentioned "read metrics", and if you mean quality scores etc..., then this should be performed before alignment with tools such as Trim_Galore! etc...