EDIT 2: Added lots of new thoughts and suggestions
As far as I understand your question, you have created a constructed language with a constructed script with own characters not in Unicode. In addition, you have created a transcription of the conscript into latin letters and digits. Now you want to typeset some texts in your conlang.
I suggest that you create a third thing: A LaTeX Input Convention for your constructed language. I suggest using macros, one macro for each character of the constructed script. This is easy to implement, and you can switch the interpretation of the macros for the conscript to the transcription and back.
This way you can typeset your conlang both in its original script and in transcription.
When you have higher programming skills, you may also write a parser for the transcription that outputs your conscript. This is a difficult task but it has been done successfully before, e.g., in the ArabTeX package.
You may also want to look into the the babel system of language packages; they offer convenient language specific shorthands to macros. As long as you do not want to include hyphenation, it shouldn't be overly difficult to read off the necessary code changes from one of the language packages with another script (e.g. Russian, Bulgarian, or Greek) to write your own conlang babel package.
Start with the documented
.ldf (language description file) format and run the
docstrip utility to get the actual package file.
EDIT: in the LuaTeX context, look at polyglossia instead of babel.