Operational dashboard
Health, security value, and latency across your Sentinel fleet
The Operational dashboard gives you a live, fleet-wide view of your Sentinel deployment. Find it in the SUPERWISE® app under Observability → Dashboards → Operational.
Use the sentinel filter to focus on specific gateways, the time range (24H / 7D / 30D) to set the window, and the Live toggle to auto-refresh. You can also share the view as a link or export it as a PDF or PNG.
Every metric is computed from anonymous metadata only — never your prompts or model responses. See Data privacy.
The dashboard is organized into three areas.
Interactions & operations
Is traffic flowing, and which gateways are handling it?
- Active sentinels and total interactions, with the trend versus the previous period.
- Interactions over time, which you can break down by provider or by gateway.
- A per-gateway table showing status, volume, and last interaction time.
Security value & outcomes
What is Sentinel actually catching?
- Passed, Remediated, and Blocked counts, each as a share of total traffic.
- Outcomes over time, stacked so you can watch the mix shift.
- Interventions by guardrail — which guardrails are doing the work.
Performance & latency
What does Sentinel cost you in latency?
- Average and P95 policy latency — the overhead Sentinel's checks add on their own.
- Average total latency — end to end, including the upstream model call.
- Latency breakdown over time (policy versus everything else), guardrail average latency, and a per-gateway table down to P95 / P99.
Updated about 20 hours ago