The sandbox is seeded with 13 hand-built incidents, each a complete RCA pipeline run with real evidence — metrics, logs, changes and hypothesis scores — not scripted narrative. Use them to see exactly how PulseServe's AI correlates evidence, ranks hypotheses, and hands off to a human at the right moments.
How to use this handbook
- Log in to the sandbox with admin@demo.io / DemoP@ss123!.
- Open any of the 13 incidents listed below, by number (e.g. INC0000001), from Incident-Mgmt in the left sidebar.
- Go to its AI Analytics tab. It opens on the AI analysis summary: the likely root cause with its confidence, impact and evidence, and the recommended next step.
- Scroll down to Analysis details and click Show steps, then open any of the four groups to see its steps.
- Follow that incident's page in this handbook for what to look for at each step and what outcome to expect.
Reading the 15 steps
The steps are grouped into four stages, each with a status pill (Done, Pending, Skipped, Failed, or Waiting for approval):
- Signals and data quality (steps 1–5): the incident, its CI and business service, the evidence window, metrics/events/logs/traces, and the evidence-quality score with any gaps.
- Correlation and hypotheses (steps 6–10): anomalies, trace-to-log correlation, changes and deployments, the timeline and failure chain, and the ranked hypotheses.
- Root cause and approval (steps 11–12): the root cause with its evidence chain and the evidence graph, then the human review — approve, reject, or modify.
- Recommendations and recovery (steps 13–15): the recommended runbook or resolution and the remediation plan, live verification of recovery, and the memory the platform keeps for next time.
PulseServe works approve-first: when a run needs review it stops after the root cause. The summary shows an "Action needed" banner with an Open review button, and step 13 reads "Waiting for approval" — the recommendation and remediation plan are only built from a root cause a person has signed off.
The 13 RCA scenarios
The module guides
Beyond the RCA pipeline, most of PulseServe's other enterprise modules are also live in this sandbox — some richly pre-seeded, some deliberately starting empty so you can build your first one.
Other demo logins
All use the same password, DemoP@ss123!. Sign out from the avatar menu to switch.
- admin@demo.io — administrator (everything, including Admin Console)
- sarah.ops@demo.io — manager (approvals, CAB votes, SLA approval, dashboards)
- agent1@demo.io to agent7@demo.io — service desk and resolver agents
- employee@demo.io — an employee, who lands on the self-service portal
Resetting the sandbox
The nightly reset (03:00 UTC) wipes and rebuilds the entire sandbox from scratch — not just the 13 incidents, but any setting, role, or record you changed too. Useful if a test run has moved something forward and you want a clean state again; don't rely on anything here persisting past the next reset.