Hi,
We’re using Intent detection, and one limitation we keep running into is that intent seems to be based only on what the requester writes.
In many cases the real intent becomes clear after the agent has asked follow-up questions, or when the agent summarizes the issue in their reply.
It would be really useful if intent could be generated based on:
- the full ticket conversation, and/or
- the agent’s messages (agent-side intent)
This would improve both routing and reporting, especially for more complex tickets where the first customer message is vague.
