The very first thing to do is look at the patent, and try to understand what it actually does. You can also provide a detailed explanation of your software workings and ask an attorney to compare that to the given patent.
In any case, it's not Custom XML in general which has been hit by i4i. It's a very specific implementation of Custom XML functionality which is problematic.
I found this nice, short explanation on [this website1:
The i4i patent describes a document format and method of encoding
where the document content — stored in
the “raw content area” — “is totally
unstructured and has no embedded
metacodes in the data stream.” The i4i
patent further states that the
document structure’s definition is
described in a separate “metacode map”
where “for each metacode applied to
the content, an entry in the metacode
map is created which describes the
metacode and gives its position.”
Whatever you do, first understand what has been patented.