Looks like its using its own proprietary hex code format (and even so, being different lengths they don't look like clean learns). Unless someone can figure it out, it's not going to help with RTI which requires Pronto-format (which is still the industry standard).

Could he perhaps tell you the raw format of the code - for example, if it's NEC format he could tell you the specifics and then you could plug that into a program like Makehex to generate Pronto format.