I need to adapt an N8N + WhatsApp solution to simulate a human response. This means sending short messages, possibly split the response into multiple parts... ideally including a "typing..." indicator (just like when a real person is replying and you see them typing).
The use case is for a funeral service, and the client specifically does not want users to feel like they’re interacting with an AI.
I'm thinking of adding some wait nodes to slow down the responses and make them feel more natural. Any other suggestions? any resource where similar behavior has been discussed? any at all?