I prefer to "own" my design front to back. If you're going to decide
you only will work on one piece, then you are just a cog in the
machine. Now in that scenario, the cogs closer to the front seem
to get all the grease. But you also have to be good at abstract
stuff, in hopping on the latest tool before it gets popular, and at
making your abstract BS seem like it's actually a source of value
rather than a tax, if you want to advance. Working the back end
you're pretty much assured of work, provided you demonstrate a
very, very low rate of screwups and decent throughput and low
need for constant coaching / correction.
None of these are known in your present situation, so the only
useful advice is "things to ask yourself".
That, or golf with your boss's boss.