So:

((A-65)*16)+(B-65)=X

If the first two bytes are "EN", as they are in your example, A=69 and B=78.
((69-65)*16)+(78-65)
(4*16)+13
64+13
77

And 77 is ASCII for "M".

CA seems to translate to space, so I guess it uses spaces for filler.

And I guess the <20> is extracted from somewhere else, unless it's one of the spaces or something. Maybe it uses different padding for different types of names. That sounds a lot like the bass-ackwards type of thing that would be used in NetBIOS.