> For the complete documentation index, see [llms.txt](https://docs.cekat.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cekat.ai/en/chat/old-analytics/ai-agent-analytics.md).

# AI Agent Analytics

The **AI Agent** page shows how your AI is performing in Cekat. Here you can see how your AI Agent responds to *conversations*, how effective it is, and how many interactions it handles without needing a human agent. You can filter the data by date and view it for each AI Agent.

#### **Here are the main metrics on this page:**

### **A. AI Message Usage**

Shown as a *line chart* with two lines: *AI credits used* and *AI messages sent*. You can also see the *Average Credit Use Per Message*, which helps you judge how efficiently you're using AI.

### **B. Agent Handoff Rate**

Shown as a *bar chart*, this is the percentage of *conversations* handed from the AI to a human agent. The lower it is, the more *conversations* the AI handles on its own. The higher it is, the more interactions need a human.

### **C. AI Accuracy Rate (in development)**

This metric will come from the *Customer Satisfaction Score (CSAT)* after the AI auto-resolves a *conversation* the customer hasn't replied to within a set time. In that case, the system sends the customer a rating link. The results will appear as a star score (up to 5 stars) along with customer reviews, to help you gauge satisfaction with your AI.

### **D. Frequently Asked Questions**

This analyzes incoming messages and identifies the questions customers ask most often. It gives you *insight* into the topics your customers care about most, with example *messages* for each.

### **E. Business Report Analysis**

**Business Report Analysis** lets you generate a comprehensive business report based on your customer conversations. The AI carries out the analysis, and you can tailor it with keywords or instructions to uncover insights such as satisfaction, complaints and common customer responses. [See how](/en/chat/old-analytics/ai-agent-analytics/business-report-analysis.md)