My name is Wijaya Adhisurya. In the past, I implemented the classic voice identification using MFCC and also the newer MPEG7, specifically ASP audio descriptor. For learning engine, I use Gaussian Mixture Modeling.
Although I never use Google AI Engine, I am no stranger to machine learning and voice identification. With some hard work, I believe I can solve your problem. I am a quick learner and can adapt to new technology easily. I hope we can at least have some conversation to see if I am the man you want or not.