Oordinary-sapphire
Alexa through Dialog Manager API & Capture Step
In the Voiceflow dev docs, you link to this tutorial for a custom integration through the Dialog Manager API: https://xavidop.me/alexa/2023-11-21-voiceflow-alexa/ The example code serves as a base to get started.
The main issue is the Voiceflow Capture step, which is most relevant for using Voiceflow with LLMs.
The tutorial suggests using the "search query" slot type in ASK. However, this only works with a carrier phrase (like "search for" or "find out"). So, it's not possible to get the full utterance of the user, or to have an intent that is always triggered to properly forward the user request to Voiceflow.
When Voiceflow still had direct integration of Alexa, this scenario did seem to work. So there must have been a trick / method used.
Alexa is one of the major voice-interaction platforms, so I think getting this to work would be beneficial.
Can you share if there is a way to still get this done, using the capture step together with Alexa?
The main issue is the Voiceflow Capture step, which is most relevant for using Voiceflow with LLMs.
The tutorial suggests using the "search query" slot type in ASK. However, this only works with a carrier phrase (like "search for" or "find out"). So, it's not possible to get the full utterance of the user, or to have an intent that is always triggered to properly forward the user request to Voiceflow.
When Voiceflow still had direct integration of Alexa, this scenario did seem to work. So there must have been a trick / method used.
Alexa is one of the major voice-interaction platforms, so I think getting this to work would be beneficial.
Can you share if there is a way to still get this done, using the capture step together with Alexa?
