Agent details
Click an agent name in any dashboard widget to open its configuration and execution history.
Dashboard Overview
Dashboard
Monitoring
The dashboard is your organization’s central command screen. It surfaces the most important operational signals so you can spot issues, track trends, and navigate to detailed views without switching between multiple pages.
What the dashboard shows at a glance
Section titled “What the dashboard shows at a glance”When you open the dashboard, you see a consolidated view of your organization’s current state:
- System status indicators at the top, showing whether core components are healthy.
- Performance snapshots with real-time metrics on agent activity.
- Anomaly and incident alerts highlighting unusual patterns that need attention.
- Helpdesk sessions listing active and recent support conversations.
- Ticket trend analytics graphing ticket volume over configurable time ranges.
System status
Section titled “System status”The status section at the top of the dashboard provides a quick health check across three dimensions:
| Indicator | What it shows |
|---|---|
| Active agents | Number of agents currently running conversations or executing tasks. A zero count may indicate a configuration issue. |
| Active deployments | Deployments that have at least one connected connector and are ready to execute tools. |
| Connected edge connectors | Edge Connector instances currently maintaining a WebSocket connection to the platform. |
Performance snapshots
Section titled “Performance snapshots”Below the status indicators, performance cards display real-time metrics on agent performance:
- Conversations in progress — the number of active chat sessions across all agents.
- Average response time — how quickly agents are responding to user messages.
- Tool execution rate — the number of tool calls being processed per minute.
- Success rate — the percentage of agent runs that complete without errors.
These metrics refresh automatically, giving you a live pulse on how your organization’s agents are performing.
Anomaly and incident feed
Section titled “Anomaly and incident feed”The incident feed surfaces alerts about unusual patterns detected across your organization:
- Elevated error rates — a sudden spike in failed tool executions or agent errors.
- Latency anomalies — response times exceeding normal thresholds.
- Connectivity drops — edge connectors going offline unexpectedly.
- Credit consumption spikes — abnormal increases in LLM token usage.
Each alert includes a timestamp, severity level, and a link to the relevant detail view for investigation.
Helpdesk sessions table
Section titled “Helpdesk sessions table”The sessions table lists active and recent support conversations with the following details:
| Column | Description |
|---|---|
| Session ID | Unique identifier for the conversation |
| Agent | The agent handling the session |
| User | The end user who initiated the conversation |
| Status | Current state: active, waiting, or completed |
| Started | When the session began |
| Duration | How long the session has been running |
| Messages | Total number of messages exchanged |
Click any row to open the full conversation transcript and see detailed tool execution logs for that session.
Ticket trend analytics
Section titled “Ticket trend analytics”The analytics section displays graphs showing ticket volume over time. You can:
- Adjust the time range — view trends over the last 24 hours, 7 days, 30 days, or a custom period.
- Filter by agent — isolate ticket creation trends to a specific agent.
- Compare periods — overlay the current period against the previous one to spot growth or regression.
- View by status — break down tickets by open, in-progress, resolved, or escalated.
These graphs help you identify peak hours, measure the impact of agent configuration changes, and plan capacity.
Navigating to drill-down views
Section titled “Navigating to drill-down views”Every section on the dashboard is a starting point for deeper investigation. Use the following navigation paths:
Deployment details
Click a deployment indicator to see connected connectors, tool assignments, and version status.
Session transcript
Click a session row to view the full conversation, tool calls, and LLM reasoning trace.
Incident details
Click an alert in the incident feed to see the full context and suggested remediation steps.
Next steps
Section titled “Next steps” Logs & Metrics View live agent runs, tool execution metrics, and LLM usage analytics.
Managing Agents Create and configure agents that power your dashboard activity.
Managing Deployments Set up deployments to run agents and tools across environments.