Evening all,
I am looking to convert one of my UI automation into NodeRed as I am making it a lot more complex and just find it easier to visualise in NodeRed however part of my automation generates a response using OpenAI then plays that via TTS. This works perfect in the UI automation but not in NodeRed and I am wondering what I am doing wrong.
Here is the automation part:
- service: conversation.process
data:
agent_id: fcd487dfde2faf132efebec47e8fceac
text: Give a random response to welcome Max home. It can be funny.
response_variable: welcomehome
- alias: Welcome home Max
service: tts.cloud_say
data:
cache: false
entity_id: media_player.downstairs_2
message: >-
"{{ welcomehome.response.speech.plain.speech }}. If you would like
me to play some music please say, hey google yes please."
In NodeRed I have created a Call service node set it to conversation process with this as the data:
{"text":"Give a random response to welcome us home. It can be funny.","agent_id":"fcd487dfde2faf132efebec47e8fceac","conversation_id":"welcomehome"}
Then another call service node for tts cloud_say with this as the data:
{"message":"{{ welcomehome.response.speech.plain.speech }}. If you would like me to play some music for you say, hey google yes please."}
However, it literally says welcomehome.response.speech.plain.speech.
Please can someone point me to what I am doing wrong!
Thanks.