The proposal is to allow Sonos Voice Control (SVC) to forward voice commands to an external system via a configurable webhook.
When a user issues a voice command containing a specific keyword or prefix (for example, “Home”, “Assistant”, or a custom wake phrase), Sonos would send the recognized text (or resolved intent) to a predefined HTTPS URL.
This URL would typically point to Home Assistant, where the command can be processed further.
At that level (Home Assistant), users and developers could:
-
Parse the incoming text or intent.
-
Route it to a custom AI/LLM or rules engine.
-
Trigger smart-home automations, scenes, or complex logic.
-
Build a local or hybrid AI assistant tailored to their own smart-home environment.
Key characteristics of the idea:
-
No access to Sonos’ internal voice engine is required.
-
Sonos retains full control over wake word detection, speech-to-text, and privacy.
-
Only explicitly opted-in commands (based on keyword or prefix) are forwarded.
-
Communication is one-way and event-based (voice → text/intent → webhook).
This would enable an integration model similar to platforms like Philips Hue, where the core system remains closed, but events are exposed to external automation platforms.
The result is a powerful extension of Sonos Voice Control into open smart-home ecosystems such as Home Assistant, without compromising security or user experience.
