Overview of AI Agent Report
In BoldDesk, the AI Agent Report is part of the broader analytics and automation suite that helps you monitor and optimise how AI agents are performing across customer interactions. The main purpose of this report is to provide insights into how BoldDesk’s AI Agent is handling customer queries, automating tasks, and contributing to overall support efficiency. This article thoroughly explores the different metrics featured in the AI Agent Report.
How to Access the AI Agent Report
-
Log in to your BoldDesk portal.
-
Navigate to Reports Module > AI Agent Performance Dashboard > Conversation Report tab.
Overview Statistics Metrics
The overview statistics widget displays the following metrics for the selected date range:
| Metric | Description |
|---|---|
| Total Conversations | The total number of conversations initiated or created. |
| Deflected Conversations | Conversations successfully handled by the AI Agent without transfer to a human (e.g., AI‑handled or the user disengaged after the AI’s response, with no human takeover). |
| Escalated Conversations | Conversations handed off from the AI Agent to a human agent (e.g., customer selected “Talk to Person” or an agent took over). |
Filtering Options
You can use filter options and save views to customise how you analyse AI Agent performance across your support operations over a preferred time period. Common filter criteria include:
-
Date Range Filter
-
AI Agent Name
-
Handling Category
-
Handling Type
a. Date Range Filter in AI Agent Reports
Selecting a time period allows you to filter and analyse AI Agent performance metrics over a specific duration. This helps you track trends, compare performance, and make data-driven decisions.
b. AI Agent
The AI Agent filter allows you to narrow down report data based on specific AI agents configured in your system. This helps you analyse performance metrics for individual AI agents—especially useful when multiple bots are deployed across different departments or channels. By selecting a specific AI agent, you can gain targeted insights into its effectiveness, usage patterns, and contribution to support operations.
c. Handling Category
The Handling Category classifies each conversation based on whether it stayed with the AI Agent or was handed off to a human.
| Handling Category | What it includes |
|---|---|
| Deflection | Conversations completed without a human takeover. This includes AI‑handled conversations and those where the user disengaged after the AI’s response (no human agent involved). |
| Escalated | Conversations handed off from the AI Agent to a human agent (e.g., customer selected “Talk to Person” or an agent took over). |
1. Deflection
Conversations completed without a human takeover. This includes cases where the AI Agent handled the request end‑to‑end, as well as conversations where the user disengages after the AI’s response and no human agent joins.
2. Escalated
Conversations handed off from the AI Agent to a human agent. This can happen when the customer selects “Talk to Person,” workflow rules route to an agent, or an agent manually takes over.
d. Handling Type
The Handling Type identifies how each conversation was processed operationally by the system and agents.
| Handling Type | When it applies |
|---|---|
| AI Agent Handled | The AI Agent completed the conversation without any human takeover. This includes user‑confirmed handling and cases where the user disengaged after the AI’s response (no human joined). Counted under Handling Category: Deflection. |
| Transfer to Human Agents | The conversation was routed from the AI Agent to a human agent (for example, the customer chose “Talk to Person” or routing rules transferred it). Counted under Handling Category: Escalated. |
| Human Agent Takeover | A human agent manually stepped in and took control of the conversation. Counted under Handling Category: Escalated. |
Conversations Widget
This table lists each conversation as a row and shows who handled it and how it progressed.
Columns:
- Conversations – Conversation subject/preview.
- Brand – The brand associated with the conversation.
- Category – The conversation category.
- AI Agent – The AI agent that handled the conversation.
- Handling Category – High‑level outcome (e.g., Deflection, Escalated).
- Handling Type – How it was handled (e.g., AI Agent Handled, Transfer to Human Agents, Human Agent Takeover).
- Created on – Timestamp for when the conversation was created.
- Details – Click View to open the conversation’s details.
Use the pagination controls at the bottom of the table to navigate through all results.
AI Agent reports cannot be filtered based on brand.
Saving View
Once you’ve applied filters within the conversation report tab of AI Agent Performance Dashboard, you can save the configuration as a custom view for future access.
-
Click Save to open the Save View dialog.
-
Enter a name for your view and choose an access scope:
-
Private – Only visible to you.
-
Any Agent – Visible to all agents.
-
Group – Visible to selected groups.
-
-
Click Save to finalize.
Your saved view will appear in the Views panel on the left side of the dashboard.
Additionally, you can;
-
Set as Favorite – Pin frequently used views to the top for quick access.
-
Set as Default – Automatically load a specific view when opening the dashboard.
-
Edit or Delete Views – Modify filters or remove views that are no longer needed.
Frequently Asked Questions (FAQs)
1) What is the AI Agent Report in BoldDesk?
A focused analytics view that shows how AI agents handle customer conversations—measuring automation impact, escalations to humans, and overall support efficiency.
2) How do I access the AI Agent Report?
Log in to BoldDesk, then go to Reports Module → AI Agent Performance Dashboard → Conversation Report tab.
3) What metrics appear in the Overview Statistics widget within the AI Agent Performance Dashboard?
- Total Conversations — total conversations initiated/created.
- Deflected Conversations — handled by the AI without human takeover (includes AI‑handled and user‑disengaged after AI reply).
- Escalated Conversations — handed off from AI to a human agent.
4) Which filters can I use on the Conversation Report tab within the AI Agent Performance Dashboard?
Use Date Range, AI Agent, Handling Category (Deflection, Escalated), and Handling Type (AI Agent Handled, Transfer to Human Agents, Human Agent Takeover).
Note: AI Agent reports cannot be filtered by brand.
5) What’s the difference between Handling Category and Handling Type?
- Handling Category is the outcome: Deflection (no human takeover) or Escalated (handed to a human).
- Handling Type specifies how: AI Agent Handled, Transfer to Human Agents, or Human Agent Takeover.
6) What does the Conversations table show within the AI Agent Performance Dashboard?
Row‑level records with columns for Conversations, Brand, Category, AI Agent, Handling Category, Handling Type, Created on, and Details (click View to open the conversation).
7) Can I save and reuse my filter setup in the AI Agent Performance Dashboard?
Yes. Click Save to create a view, choose access (Private, Any Agent, or Group), then save. You can Set as Favorite, Set as Default, and Edit/Delete saved views later.