Is there an explanation/example of how speech interpretation is handled using this device - such as accessing external recognition processing for phrase recognition? I image integration with Siri is not an option unless the app is about ride sharing or another of the limited domains available to developers. Or does this function based on specific words or phrases to recognize?
We only produce the hardware, though we want to create guides on how to get different services running.
Right now here are some options.
Online: Alexa, Google Assistant, DuerOS (Chinese).
Offline: PocketSphinx, and hopefully Snips.ai soon.
There are additional options you can find by searching online as well, though we may not have tested them. You could try to configure them yourself, and if there is enough interest we can help configure them as well.