Hey all,

I am working on building a fully capable AI personal assistant that is 100% local. It is going to be a self evolving, Learning AI assistant that will integrate with things like Home assistant. I have it mostly built, still working on testing and getting satellite speakers and displays to work. it is built using the Qwen Family. However, it does not rely 100% on the LLm, there is a 3 layer architecture that essentially captures the intents and will direct things as it comes in with the LLm being the last fall through.

this is the blurb I have " ... transforms a local LLM into an intelligent home assistant that understands who's speaking, adapts to each family member, controls your smart home, and gets smarter every day — all running on your hardware, with zero cloud dependencies."

The question I have, I want to train a new wake word (I know how to train it) but I need actual audio samples of people saying the wake word. Does anyone know of a good place to crowd source people saying it?

Thanks in advance?

btw: I didn't post the link to the repo because right now, I am not trying to self promote even though It is going to be fully open source. If this is something of interest, I can post it, it just is not ready yet.