HelixGPT Performance and Token Consumption Dashboard

As an administrator, use the HelixGPT Performance and Token Consumption Dashboard to monitor, analyze, and optimize HelixGPT's operational behavior across your environment.

The dashboard provides detailed visibility into key performance metrics, including total conversations, token consumption patterns, document access, AI agent‑level activity, and latency trends.

By consolidating performance and usage data into a single view, the dashboard helps you identify high‑usage AI agents, evaluate response efficiency, and assess the overall impact of BMC HelixGPT workloads. These insights support effective capacity planning, cost management, and continuous performance tuning.

The following image shows the HelixGPT Performance and Token Consumption Dashboard with sample data:

Before you begin

To access the dashboards, make sure that BMC Helix Innovation Suite is deployed in your environment. 

To view the HelixGPT Performance and Token Consumption Dashboard

1. Log in to BMC Helix Dashboards.
2. Expand the left navigation pane, and click Dashboards.
3. Search for the HelixGPT folder and click it.
4. Click HelixGPT Performance and Token Consumption Dashboard.
5. Select a time range for data. 
   The data for the selected time range is displayed.
   By default, the dashboard displays data for the last 30 days.

HelixGPT Performance and Token Consumption Dashboard filters

You can look for specific information on the dashboard by using the following filters:

Filter	Description
Time range	Filter by the time range, such as Last 1 hour, Last 30 days, or Last 1 year.
Country	Filter by the country name where the BMC HelixGPT application is used.
Organization	Filter by the organization that uses the BMC HelixGPT application.
Source channel	Filter by the name of the application that has BMC HelixGPT enabled.
Source Agent	Filter by the BMC HelixGPT agent used in the application.

Panels in the HelixGPT Performance and Token Consumption Dashboard

The following table describes the HelixGPT Performance and Token Consumption Dashboard panels: 

Panels	Description	Example
Total HelixGPT Conversations	Displays the total number of unique BMC HelixGPT chat sessions.
Estimated Total HelixGPT Token Usage	Displays the number of total (input + output) tokens used through all the chat sessions.
Estimated HelixGPT Input Token Usage	Displays the total number of input tokens used across all chat sessions.
Estimated HelixGPT Output Token Usage	Displays the total number of output tokens used across all the chat sessions.
HelixGPT Document Usage Count	Displays the total number of documents referenced by all AI agents across all chat sessions.
HelixGPT Interactions by Source Agent	Displays a list of AI agents used to answer user queries. The panel also displays the number of times an agent was used.
HelixGPT Estimated Token Usage by Source Agent	Displays a list of AI agents that used the most tokens to answer user queries. The panel also displays the number of tokens each AI agent used.
HelixGPT Average Latency by Source Agent	Displays the list of AI agents along with the average time each agent takes to respond to user queries. The panel also shows the average response latency (in milliseconds) for every AI agent.
HelixGPT Document Count by Agent	Displays the list of AI agents that referenced the most documents to answer user queries. The panel also displays the total number of documents referenced by each AI agent.
HelixGPT Token Usage by Agent (Monthly)	Displays the number of tokens each agent used during the selected month. The panel also includes a summary table listing the mean token usage per agent.
Total HelixGPT Conversations vs. Avg Latency (Daily)	Displays the comparison of the daily number of HelixGPT conversations with average response latency.