Hi,
I assume the OH is not from carboxylic acid and so you don't want the amide. I dont know to couple them directly and I'm not sure how is it doable with CDI. I think one would rather take the amine to hydroxylamine and convert the OH to a good leaving group. There is a procedure: triphenylphosphine, carbon tetrachloride, and DBU in the presence of catalytic amounts of TBAI in acetonitrile to get the oxime (R1R2C=NOH and alcohol R3OH) to R1R2C=N-O-R3.
Other ways to place an N-O bond that I can think of at the moment:
-Mitsunobu with hydroxyphthalimide on the alcohol (OMs, OTs...)
-Mesylate, then DBU/Boc-N-OR, deprotect
-There is some Buchwald chemistry which allows N-O formation
-Copper-catalyzed N-arylation of hydroxylamines if you've got N-Ar
Not sure if this is of any use;) good luck, maybe I'm missing something simple.