You could do something that exploits the greater acidity of the O-H relative to NH2. Perhaps one equivalent of NaH to deprotonate (this can be tricky in the real world because NaH is not so soluble so maybe use a different base) and then functionalize the alkoxide into a tosylate, eliminate with bulky base, then hydrogenate.
I am not sure if deprotonating first is even necessarsry, but I am not sure if the NH2 would too nucleophilic to selectively tosylate an alcohol in its presence. Making an alkoxide I figured would solve this problem.
This is the "sophomore" level organic chem. There is probably a better way to do it in real life if that's what you're going for. Search for "remove primary hydroxyl in the presence of primary amine" or something like that on sci finder, or even do a structure search.