Anyone have some insight and willing to share what technologies are used to build voice assistants such as RASA[1], Alexa, Siri, etc?. How do they structure their platforms and what is the architecture behind it.

[1]: https://rasa.com/docs/