There are two ways to input the voice assistant Alexa to different hardware.
One is purely soft way. I put the voice assistant for the echo sound system, no matter who wants to use it just put it into the software. This type of technology at a maximum need to probe into the speech recognition, the core is the semantics and dialogue. The type is obvious refer to search engine.
The other one is the way soft and hard to integrate for Amazon Alexa sound smart speaker . I make a standard hardware product like smart Bluetooth speaker, and then open the assistant Alexa in the hardware wireless Bluetooth speaker products. The most obvious reference to this type of operating system is Android.
The pure internet company will prone to the first way as it is lightest. And for the later it need to do the manufacture and sales, and touch the acoustics. Undoubtedly, it is heavier.