What do you want to achieve?
Disable agent responses for specific customers.
What issues are you encountering?
Malicious customers keep sending messages, causing the AI agent to continuously consume information.
Contact profile + department, combined solution.
Related tutorial: 以"部署"功能实现灰度上线功能
Strategy 2 in
We also encounter similar issues. You can refer to our practices:
-
Post-incident detection and handling: When identified, update the custom fields for the relevant contact, such as setting “Malicious Customer” to “Yes”. Then, in the Agent [Deployment] menu, configure the system to only respond if the contact is not marked as “Yes” for malicious customer.
-
Earlier intervention: In the [Guidance] section, under the conditions for transferring to human agents, use a prompt to specify that if there is no progress toward business objectives for three consecutive turns (you can specify your company’s specific goals), or if the customer’s questions are not answered, then transfer to a human agent. Once the transfer to human agent is triggered, the 3Chat backend will send an email/SMS notification to the administrator/human agent, and the AI will pause and stop responding.
Option 2 is not quite suitable because we must configure the AI to come back online after a fixed interval once the conversation is transferred to a human agent.
Option 1 is difficult to implement. I’ll give it a try based on the instructions. Thanks for the reply.
I’ve already set it up according to the steps. However, the field content on the Contacts page cannot be edited directly. Do I need to use another tool to make changes?
Both of these locations can be modified.