During testing, the backend indicates a manual reply, but no one is logged into WeCom. As a result, no one replies to messages, yet the system repeatedly follows up asking if the user is still there. What could be causing this?
Let’s clarify one point first: does “no one is logged into WeCom” mean there are no managed accounts?
Once manual mode is enabled, the AI will no longer respond; only manual responses will be possible.
In the managed account, when I tested communication via WeChat with this managed account, it immediately indicated human assistance. Therefore, when I sent a WeChat message, there was no response. However, after a period of time, once the conversation automatically closed, the agent came online to continue the dialogue. We never specifically activated the human assistance mode.
No, this is a setting on the Enterprise WeChat side for a unified greeting message. It has always been like this before.
Can you share a screenshot of the conditions for transferring to a human agent? Have other contacts experienced this, where customers are transferred to a human agent immediately upon arrival?
The issue has been identified and resolved in the latest version iteration.
The cause was a timing mismatch between the initial greeting on WeCom and the AI’s response. Please continue to monitor the situation, and if any issues arise, feel free to leave a comment.
May I ask why the AI keeps replying to transfer to a human agent, but the transfer actually doesn’t happen?
Please click this icon; it will display “AI Execution Process.” Then, upload a screenshot of it so we can investigate the issue.
