The current Human in the Loop feature is excellent for allowing users to request human assistance, but I'd like to propose an enhancement that would allow administrators/agents to proactively take over conversations without requiring user initiation.
Currently, human intervention can only occur when triggered by the user, which limits our ability to provide timely assistance when we observe issues in ongoing conversations. As bot builders, we often monitor conversations and notice when the bot is struggling or misunderstanding a user, but we can't intervene directly.
I'd like to request the ability for authorized agents to:
- Monitor live conversations from the back office > done already
- Initiate a takeover at any point
- Seamlessly enter the conversation to provide assistance
- Hand control back to the bot when finished
This would significantly improve customer experience by allowing us to catch and resolve issues before they become problematic, particularly for complex bots where edge cases are difficult to anticipate.
Thank you for considering this enhancement!