hacking on tex parsing

Hey folks,

I'm hoping someone can point me in the right direction here -- I would like
to make some TTS software to help me proofread my papers, which are all tex
(I use pdflatex to compile). I know very little about how pdflatex works,
but I'm hoping that I can reuse some existing tex parsing code. So for
instance, if I can grab the document after newcommand or
DeclareMathOperator has been processed that would be very helpful. Or if
there's some tree-like data structure that gets created that kind of thing
could be useful too.

I know there are things like TexSoup, but those only parse the tex source,
they don't perform any of the processing operations I'd like to have.

Thanks!