App Usage Analytics Dashboard¶

The Analytics Dashboard provides real-time insights into app usage with comprehensive visibility into user engagement, performance metrics, and resource utilization for agents, tools, and models.

The dashboard organizes analytics into two primary categories:

Session-level insights: User activity and engagement metrics.
Run-level insights: Execution performance and resource consumption.

Key Features¶

Real-time data updates.
Customizable time frame and environment filters.
Trend analysis with historical comparisons.
Interactive visualizations with detailed drill-down capabilities.
The default view displays the current draft environment.

Customize the analytics view by using the dashboard's filtering options and interactive controls:

Time Frame Selection: Choose from pre-defined ranges or set custom dates to analyze specific periods.
Environment Filter: Select specific environments for targeted analysis (defaults to current draft).
Environment Filter Label: Displays a visible label for the Environment filter to improve discoverability.
View Modes: Toggle between interactive charts for trend analysis and detailed tables for granular inspection.
Drill-down Navigation: Click any metric card or data point to access expanded insights and detailed breakdowns.
Export Options: Download data for offline analysis and custom reporting.

Key Metrics and Components¶

Session Metrics¶

The dashboard's top section, displays four key performance indicators, each featuring:

Current value with real-time updates
Percentage change from the previous period
Visual trend indicators
Interactive time-series charts

Users¶

Tracks unique active users within the selected timeframe, displaying:

Total user count
Period-over-period comparison
Daily or hourly activity breakdown

Sessions¶

Monitors total session volume, including:

Session count and trends
Historical performance comparison
Time-based distribution chart

Messages¶

Measures communication volume between users and the system:

Total input and output message counts
Comparative trend analysis
Hourly and daily volume breakdowns

Tokens¶

Tracks token consumption across Agent and Supervisor components:

Total tokens consumed
Usage comparison with previous periods
Time-based consumption patterns

Run Analytics¶

The Runs section provides detailed execution metrics across three component categories, available in both chart and tabular formats.

Agent Runs¶

Monitors agent execution performance with:

Total execution count and trends
Performance metrics over time
Detailed agent-level data: Agent name, Number of runs, Average response time, and Token consumption

Tool Runs¶

Analyzes tool utilization across all agents:

Execution counts with trend indicators
Breakdown by tool type: Workflow, Code, MCP (Model Context Protocol), and Knowledge
Tool-specific metrics: Tool name, Run frequency, Average response time, and Tool category

Model Runs¶

Tracks model API calls from Agents and Supervisors:

Total model invocations
Performance trends over time
Model-level details: Model name, Execution count, Average response time, and Token usage

Send Feedback