Click , select an impacted user from the User List table and click
AI Assist to open the co-pilot sidebar.
Formulate and submit your troubleshooting query. Do one of the following:
- Enter your question in natural language, for example, "Why is Jonsmith
having a degraded experience on figma.com?".
- Select a pre-suggested prompt, such as "Diagnose why Susan Jobs is
having a bad experience on Microsoft Teams.". Select the query and click
Modify to update the query before you execute
it.
Review and execute the diagnostic plan.
- The AI Assist presents a detailed diagnostic plan outlining
investigation steps, such as assessing device health, reviewing
real-user experience, and performing synthetic tests.
- Click the expand button to view the co-pilot window in full screen.
- To refine your initial prompt before execution, click
Modify
- To proceed with the analysis, click Execute.
Monitor the execution and view the final analysis report.
- Observe the Copilot sidebar as it displays intermediate updates and
findings during the execution phase.
- Upon completion, review the Executive Report, which includes:
- Root cause determination - AI Assist correlates findings
across all domains to identify the true root cause rather than
just the most visible symptom. For example, critically high
memory usage on the user's client device.
- Supporting evidence - Shows specific data points and
metrics to support the root cause identified.
- Blast radius analysis - The system determines whether the
issue is isolated to a single user or affects multiple users,
helping you prioritize remediation efforts.
- Timeline and baseline comparison
- Diagnostic journey summary - AI Assist analyzes the issue
and creates a diagnostic plan tailored to the specific user and
application context.
- Recommendations for immediate and long-term actions.
Based on the
Recommendations section of the report, take the suggested
steps to address the root cause.
