Customer Question Learning

——Learn how to operate AI Agent knowledge bases through this article, improving knowledge operation efficiency.

The Role of Customer Question Learning

We hope that before you learn about the [Customer Question Learning] feature, you understand its scenarios and uses:

● Usage scenario: When AI Agent faces high-frequency questions or businesses, the enterprise's knowledge operators find customer questions and make optimizations.

● Purpose: To cluster high-frequency customer questions. Operation staff can accurately identify the robot's response weaknesses based on metrics such as problem resolution rate and trans-to-agent rate. This allows for targeted optimization of the knowledge base and response strategy, improving knowledge operation efficiency and the robot's independent reception rate.

How to Use Customer Question Learning

The following will introduce you to the function and effect of each feature point:

● Data Filtering Rules and Conditions

1. Supports filtering by time, with the smallest unit being days.

2. Support filtering by robot.

3. Support filtering by business type.

4. Support filtering by channel type.

5. Support filtering by whether the robot triggers a trans-to-agent process.

6. Support filtering by question language.

7. Support filtering by resolution status.

8. Support filtering by response rating.

9. Support for searching by entering questions.

● Customer Clustering Question Sources

Introducing the source of the customer clustering problem in customer question learning for you:

When customers consult with the robot, each original customer question will be clustered into the formed categorized questions.

● When to Use Customer Question Learning

1. Start Time: Once the robot is online for official user use, you can start using customer question learning. The learning frequency depends on the data volume.

2. Continuous operation time: When the robot's independent reception rate shows no clear increase or decrease during the observation period, you can reduce the frequency of learning. Checking once a week is sufficient.

● How to Use Customer Question Learning

1. On the customer question learning page, you can see the data filtering rules and conditions. We can perform filtering or manually input to search for question.

Figure 1： Customer Question Learning Page

2. Click [Details] to view the specific customer issues and other information about the clustering problem.

Figure 2： Detail Page of The Clustering Question

3. Click [Chat Details] to display the conversation information, which provides information for adding knowledge to clustering problems.

Figure 3：Chat Details Page for Customer Question

4. Click [Add Knowledge] to optimize the knowledge base.

Figure 4：Add a Knowledge Page for Customer Question

● Other Features

1. Support exporting reports.

2. Support custom filter fields.