I think we need to consider the need as a whole 
Before talking about development, we should create a small functional specification: what do we want to do?
How does it work at competitors?
For example, at Alexa, Google Home, when a question is asked, « who » in the house is responsible for the action?
After that, with Google / Alexa, itâs simply you speak, and it executes what you ask. 
But I can give you a quick spec
On Alexa, there are voice profiles, so it can speak to different people in the house and we can customize the responses, I think
Yes, but some commands require knowing who is speaking, right? If you say « Add XXX to my shopping list » or « Tell YY that Iâm leaving the house, » what happens?
I think itâs worth looking at how they handle this and thinking carefully about the usage
Similarly, the Rhasspy part, it would be great to define what we want in terms of interface in Gladys
Yeah but anyone can talk to her, she executes
I agree thatâs the desired functionality in principle.
And the second issue is that we donât all have Amazon/Google engineers to do profile recognition for us ^^
We can consider planning for a case where we recognize who is speaking. But it wonât work right away, as far as I know, no open-source tool is currently capable of doing this.
However, yes, implementing this possibility at the code level can be considered.
We donât execute themâŠ
Of course! Iâm not saying we should do it, we just need to plan ahead and make sure Gladys doesnât crash if a request requires knowing who is speaking.
Also, we will need to block requests that have non-textual responses: « show me the living room camera » is of little interest vocally.
Yes yes of course!
My 4-year-old daughter, if you want a crash test (she makes Alexa blow up ^^)
My niece asked to play « soupe aux choux » because she had that song at school, Alexa played « bite au chaud » ⊠(4 years old) lol
My son is 2 and a half years old, itâs going to be great!!
He already touches all the buttons he finds and uses the touch screen on the TV.
My 4-year-old daughter made it crash too 
Hello hello,
I came across a great offline and open-source voice recognition solution (on a Discord server)
It really works like a charm. I tried the small model (39MB) and it works perfectly and in real-time.
Plus, the documentation is well done and it installs in 2 minutes.
Honestly, I think itâs a great lead for gladys 
(And otherwise, nothing to do with it but do you know how to change your username on the forum 
?)
In your account preferences, I think? You click on your profile and go to preferences
If itâs not possible, send me a PM with your new username and Iâll change it.
I donât know if this is still up-to-date, but I came across this link a few days ago: User has no option to change username? - #6 par pfaffman - Support - Discourse Meta
To keep in mind, therefore
No, we canât, but @Toulouse-wine changed it for me, thank you! 
Hello, do you have any news or tests regarding voice? Iâm waiting for this to change version 
Hello,
Not much, actually moving + a little girl who arrived earlier than expected
As a result, I couldnât make progress on my integrations⊠For the moment, itâs at a standstill, but I hope to be able to resume soon.