question

I'd like to see a way to be able to dynamically supply a list of possible choices for any given slot and have the API phonetically find the closest match. My current flow is: * User: "tell X to turn bob's office lights on" * App: only "office" is returned, so the app asks user which lights and lists all possible choices * User: "bob's office" * App: only receives "office", and thus prompts again and the user is unable to achieve their intent The only way to make this work currently is to have "bob's office" in the sample utterances file, which obviously is not a realistic option. The other option is to be able to receive the full text of what Alexa heard, and do the choice logic ourselves.