I’ve noticed that the AI consultant directly addresses clients by their WeChat names during conversations. Is it possible to disable the automatic retrieval of WeChat names and instead use generic greetings such as “Hello,” “Sir/Madam,” or “Teacher” until the client specifies how they wish to be addressed? I couldn’t find where to adjust this setting. Some WeChat names are quite unusual, and using them as forms of address can be awkward and overly robotic.
Hello, this can be constrained in the Agent’s condition following settings by requiring that the response does not include the client’s name, using only generic greetings like “Hello,” “Teacher,” etc.
By default, without sufficient conversation data, the AI cannot determine whether the client is male or female. Therefore, it is recommended not to use specific gendered titles like “Sir” or “Madam.”
I encountered the same issue, and here’s how I resolved it:
Try adding the sentence shown in the image (this approach has been proven to work). You can also adapt it slightly depending on the industry. For example: “Baozi” (darling), “classmate,” “this parent,” or “sister.”
To summarize, add the following sentence to the “communication style” section in the image (this hasn’t been tested yet, so you’ll need to try it out):
At all times, it is forbidden to address users by their names. You are only allowed to use the most appropriate term from the following list: [“You,” “Baozi,” “classmate,” “this parent,” “sister”].