It seems there are several services already called "Vocafy" - plus one called "Vocalfy" - so you may want to reconsider the current name if you want this to become popular.
I've been working on a project to implement a vocal synthesizer in C. Specifically, based on Dr. Jordi Bonada's PhD thesis "Voice Processing and Synthesis by Performance Sampling and Spectral Models". In part of that paper, he discussed his "Wide-Band Voice Pulse Modeling" algorithm. In this post, I've presented a few ideas I've had to improve the algorithm (based on the flaws he noted), and some early test results of my implementation of them.