> For the complete documentation index, see [llms.txt](https://alhena.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://alhena.gitbook.io/docs/dashboard/analytics/agent-performance.md).
# Agent Performance
Agent Performance Analytics
#### 1. Where to find it
1. **Open the Alhena Dashboard.**
2. In the **left-hand navigation**, click **Analytics**.
3. Across the top of the Analytics workspace, choose the **"Agent Performance"** tab
***
#### 2. What you see at a glance
| Metric card | What it tells you |
| ------------------------------------------- | ------------------------------------------------------------------------------------------------ |
| **Total conversations** | Total number of conversations across all channels during the selected time period. |
| **AI profile conversations** | Number of conversations handled by your AI agent. |
| **Human agent conversations** | Number of conversations that were handled by human agents. |
| **Resolution Rate** | Percentage of conversations resolved by AI out of total conversations. |
| **Agent Assist conversations** | Number of conversations where Agent Assist was used. |
| **Credit usage** | Total credits consumed across conversations (visible depending on your plan). |
| **CSAT** | Average customer-satisfaction rating for the period. *(Website, Nudge, and Icebreaker sources.)* |
| **Conversations started by tapping AI FAQ** | Conversations a visitor began by tapping an AI-generated FAQ. *(Product FAQ source.)* |
| **Conversations started by direct message** | Conversations a visitor began by typing a message directly. *(Product FAQ source.)* |
Each card is color-coded to match its data series in the chart below, so you can cross-reference quickly. **The metrics displayed vary based on your selected integration source** — for example, CSAT appears for Website/Nudge/Icebreaker, the two FAQ breakdowns appear for Product FAQ, and social channels show a smaller subset. Credit usage is only shown on credits-based plans.
{% hint style="info" %}
Each metric card also shows a **percentage change** indicator. This percentage compares the current period with the equivalent previous period. For example, if you select **Last 30 days**, the percentage compares those 30 days against the 30 days before that. Similarly, selecting **Last 60 days** compares with the preceding 60-day window.
{% endhint %}
For the **Website**, **Product FAQ**, **Nudge**, and **Icebreaker** sources, two additional breakdown panels appear below the chart:
* **Top 10 countries** — the countries driving the most conversations for the selected source.
* **Top 10 URLs** — the pages driving the most conversations for the selected source.
***
#### 3. How metrics are calculated
**Total conversations**\
The sum of all conversations — both AI-handled and human-handled — during the selected time period.
> **Total conversations** = AI profile conversations + Human agent conversations
**AI profile conversations**\
Conversations that were fully resolved and closed by the AI agent without any human intervention.
**Human agent conversations**\
Conversations that were transferred to a human agent. This includes cases where the AI could not resolve the query and escalated it, or where the customer explicitly requested a human agent.
**Resolution Rate**\
The percentage of total conversations that were resolved by the AI agent.
> **Resolution Rate** = (AI profile conversations ÷ Total conversations) × 100
For example, if there are 800 AI profile conversations out of 1,000 total conversations, the resolution rate is **80%**.
**Credit usage**\
The number of credits consumed for AI profile conversations that were resolved by the AI agent. Credit usage is only visible for accounts on a credits-based plan.
***
#### 4. Global filters
| Control | Why it matters |
| --------------------------------------------------- | ------------------------------------------------------------------------------------------- |
| **Integration source** | Filter by channel (Website, Email, Slack, Discord, etc.) to see performance by integration. |
| **AI Profile filter** (default **All AI profiles**) | Compare performance across multiple AI profiles or focus on a specific one. |
| **Date range** (default **Last 30 days**) | Supports presets (Last 7, 15, 30, 60 days) and custom ranges. |
Filters automatically refresh both the metric cards and the underlying chart.
***
#### 5. Practical use-cases
| Goal | How to use Agent Performance Analytics |
| ------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- |
| **Evaluate AI training improvements** | Compare AI profile conversation counts and resolution rate before and after updating your training data or guidelines. |
| **Identify peak traffic periods** | Use total conversation trends to spot busy periods and ensure adequate human backup coverage. |
| **Reduce human workload** | Monitor the resolution rate—a higher rate means more conversations are being fully resolved by AI without human involvement. |
| **Track Agent Assist adoption** | See how often your team uses Agent Assist to help with responses. |
| **Monitor costs** | Track credit usage over time to understand consumption patterns and optimize your plan. |
| **Benchmark period-over-period** | Use the percentage change indicators to compare performance against the previous equivalent period. |
***
#### 6. FAQs
**Why do I see different metrics for different integrations?**\
Metrics vary by channel—filter by a specific integration to see metrics relevant to that source.
**Why don't I see credit usage?**\
Credit usage is only visible for accounts on a credits-based plan.
**What is Agent Assist?**\
Agent Assist helps human agents by suggesting AI-generated responses during conversations.
**How is the percentage change on each card calculated?**\
The percentage compares the selected time period against the immediately preceding period of the same length. For example, selecting "Last 30 days" compares with the 30 days before that.
**What counts as a "resolved" conversation?**\
A conversation is considered resolved when the AI agent fully addresses the customer's query and the conversation is closed without being transferred to a human agent.
***
**Agent Performance Analytics gives you clear visibility into how effectively your AI is serving customers—helping you track conversation volume and measure the balance between AI and human-handled interactions.**
***
#### See also
* [Trending Topics](/docs/dashboard/analytics/trending-topics.md) - Analyze conversation patterns by topic to understand what customers are asking about
* [Revenue Impact](/docs/dashboard/analytics/revenue-impact.md) - Measure the revenue contribution of your AI agent with cart and GMV analytics
* [CSAT Collection](/docs/dashboard/analytics/csat-collection.md) - Set up customer satisfaction feedback collection after AI interactions
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://alhena.gitbook.io/docs/dashboard/analytics/agent-performance.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.