I have an idea for you, but I don't know if you can realize it. For example, we have VSL samples of a voice or of several voices in the choir singing o,a,u vowels. I imagine a software with artificial intelligence that adds the desired consonants to exactly these vowels. For example, A becomes "Adam", and any words are then created from all three vowels.
In addition, I have been working with computer art since 1992, where I also won prizes. These activities I always practiced as a sideline, mostly as a hobby. Currently I have started to work in the field of art as an AI operator.