It would be great to have an option to contribute successfully captured voice instructions to Mozilla Voice directly from HA. This way the future STT/TTS models can be improved further.
I believe Mozilla Voice is a custom integration or add-on, right? I’ve not heard of it.
If that is the case, please contact the authors to extend the software, not HA…
There is no HA addon/integration for Mozilla Voice that I know of. Actually, the correct and full name is Mozilla Common Voice (Common Voice) and it’s a project to collect freely available data that can be used to train models for speech recognition. Having a quality model means that HA will be able to understand voice commands better.
I think it would be nice to give HA users an option to contribute their sentences to the voice project hence improving the experience for everyone else.
Not really a Feature request, then.
This link is relevant to your request found in the Pinned Post for this this category.