You are here:
Fulfiller Resolves an Incident with Proactive Assistance
Help an incident fulfiller understand an issue, evaluate an AI-proposed fix, and restore service with a verified resolution.
Use Case at a Glance
Persona: An incident fulfiller responsible for restoring an employee's IT service
Job to be done: Resolve an assigned incident efficiently without overlooking record context, related incidents, or proven solutions
Starting point: An open incident in the Agentic IT Service Desk with the Proactive Assistance card
Successful outcome: The fulfiller applies and verifies an appropriate fix, records the resolution, and sets the incident status to Resolved.
This use case connects Proactive Assistance insights to a fulfiller's resolution decision. It doesn't replace the canonical instructions for configuring actions, using the card, or managing incidents. Instead, it shows how the fulfiller reviews AI-generated context, validates a proposed fix, and retains responsibility for the final resolution.
Start with AI-Prepared Context
When a new incident is created, the following actions can run automatically: Summarize Incident, Associate Similar Incidents, and Check Incident Attributes. Check Incident Attributes compares category, subcategory, urgency, and impact with similar past incidents. The action can automatically update the category—for example, to VPN—and display a message card.
Propose Resolution Summary runs when the fulfiller is ready to resolve.
Create Incident Root Cause Summary runs after the incident is resolved. Create Service Plan is optional when the incident needs a multi-step task plan in Service Assistant.
Use these insights to understand the issue before changing the record or applying a fix. Confirm that the summary matches the employee's report, related incidents reflect the same symptoms, and the incident attributes support the appropriate routing and priority. For action behavior and examples, see Proactive Actions for IT Incidents.
Proactive Assistance accelerates triage, but its output is evidence rather than a final decision. If the card reports insufficient data or the suggested context doesn't match the incident, investigate the record directly instead of treating the generated output as authoritative.
Turn Insights into a Resolution Decision
Review the signals together to choose the next action instead of accepting every suggestion or using every proactive action.
| Signal | Fulfiller Check | Resolution Decision |
|---|---|---|
| Incident summary | Does it reflect the reported symptoms, impact, and latest updates? | Continue with the summarized context or review the complete record history. |
| Similar incidents | Do the linked records describe the same issue and environment? | Reuse a proven approach or separate unrelated records. |
| Check Incident Attributes | Do the category, subcategory, urgency, and impact values on the incident record match the evidence? | Keep the automatically updated values when they match the evidence, or correct the incident before continuing. |
| Create Root-cause summary | Does the proposed cause align with related incidents, known errors, or recent changes? | Use it to guide investigation or gather more evidence. |
| Proposed resolution | Is the fix relevant, safe, and supported by historical resolution notes or Knowledge? | Test the fix, refresh for another proposal, or continue manual investigation. |
Propose Resolution Summary analyzes historical resolution notes and relevant Knowledge articles to suggest steps. The fulfiller evaluates the recommendation and executes the appropriate work; Proactive Assistance doesn't autonomously remediate or close the incident. For the full triage-to-resolution model, see Proactive Assistance Workflow for IT Service Records.
Apply and Verify the Fix
Before making a change, consider its scope and risk. A routine, reversible fix can proceed through the team's standard operating practice. A fix that changes shared infrastructure or introduces broader risk belongs in the appropriate change-management path.
Apply the selected fix and verify the outcome against the original symptoms. Confirm with the affected employee when the resolution requires user validation, and check related monitoring or records when the issue has a wider scope. If the first proposal doesn't solve the issue, refresh Propose Resolution Summary for another suggestion or continue the investigation with the evidence already gathered.
If the incident needs a series of tasks, the Create Service Plan action creates a step-by-step plan. Fulfillers find that plan in the Service Assistant component, not on the Proactive Assistance card. Refresh the action if incident details change. Keep the use case focused on the resolution outcome; see Work with Service Assistant for IT Incident Resolution for plan execution and lifecycle details.
Record a Complete Resolution
After the fix is verified, record what failed, what the fulfiller changed, and how the result was confirmed. A useful resolution summary gives future fulfillers reliable evidence for similar incidents and improves the historical data that Proactive Assistance uses.
Set Status to Resolved when the issue is fixed, enter the resolution details in Resolution Summary, and save the record. Use Closed only when the incident lifecycle is complete under the team's process. For the canonical status behavior and incident-management procedure, see Manage Incidents for IT Services.
The adoption milestone occurs when a fulfiller uses proactive context to reach a sound decision, verifies the fix, and leaves a complete resolution record. The value is faster understanding and reuse of proven evidence, not automatic acceptance of an AI-generated answer.
Confirm That Proactive Assistance Is Ready
Before using this journey as an adoption test, confirm that Proactive Assistance is configured for incidents and that the fulfiller has the required access. The feature depends on correctly configured Data 360 data and search indexes. See Set Up Proactive Assistance for IT Service Records for setup and permission requirements.
Missing suggestions aren't proof that no relevant solution exists. Recommendations can be unavailable when Data 360 data or search indexes are not ready, records have not synced, or there is not enough similar history. Missing suggestions are not proof that no solution exists. Investigate the record directly. See Considerations and Limitations for Agentforce in IT Service for data and retrieval limitations.
