You are here:
Admin Validates the First Agentforce IT Service Incident
Validate a focused Agentforce IT Service deployment with one employee issue. Confirm that portal intake, assignment, routing, and verified resolution work together before expanding the deployment.
Use Case at a Glance
Persona: A Salesforce admin responsible for validating the first Agentforce IT Service deployment
Job to be done: Confirm that the configured employee and fulfiller experiences work together across one complete incident journey
Starting point: The required Agentforce IT Service capabilities, data foundation, Employee Services portal, pilot users, Knowledge, assignment, and routing are ready
Successful outcome: A pilot employee creates an incident without admin assistance, and the intended fulfiller receives and resolves it. The employee can retrieve the updated status or receive a configured in-app or channel notification.
Use this scenario to turn initial setup into a validated first-value milestone. Define a controlled pilot, connect the responsible components, and evaluate whether the first incident journey is ready to expand. See the linked Help topics for procedures to deploy Agentforce IT Service, configure agents and the portal, manage access, and route incidents.
Confirm Agentforce IT Service Readiness
Before using this journey as an adoption test, confirm that the selected Agentforce IT Service capabilities are deployed and that Einstein Generative AI and Agentforce are on. Confirm that Data 360 is on, that the IT Service Data Kit is deployed, and that the required Knowledge search index is available. Also confirm that the IT Service Employee agent has the required subagents, actions, access, and portal deployment. If the pilot includes assistance from the IT Service Fulfiller agent, confirm its configuration and access separately. See Set Up Agentforce IT Service Using Setup Bundles, Prerequisites to Set Up Agentforce for IT Services, Install the IT Service Data Kit, and Create a Knowledge Search Index for Agentforce IT Service.
Confirm role-appropriate access for the admin, pilot employee, and pilot fulfiller. A profile alone doesn't provide every required Agentforce IT Service permission. See Permission Sets for Agentforce IT Service and Agentforce Permission Sets for IT Fulfillers and Employees.
Allow for data sync and indexing before diagnosing recently created records. Recently created incidents may not be retrievable by record-identification actions until search indexes catch up. Confirm the incident record itself first, then retest retrieval after indexing completes. An immediate failure to find, summarize, or comment on the record doesn't prove that incident creation failed. Validate the incident record directly, and then retest retrieval after the indexing window when record-identification actions are part of the pilot.
Define the First-Incident Pilot
Choose one frequent, low-risk employee issue that a pilot fulfiller can reproduce and resolve. Match the issue to the configured IT Service Employee agent behavior, and provide enough information for assignment. Select an issue with a known resolution for the fulfiller to verify.
| Pilot Decision | Admin Question | Evidence of Readiness |
|---|---|---|
| Test issue | Is the issue specific, repeatable, and safe to resolve in the pilot environment? | The employee, fulfiller, expected fix, and verification method are identified. |
| Employee entry point | Can the pilot employee use the deployed IT Service Employee agent in the Employee Services portal? | The employee logs in with their own credentials and can open the agent without admin assistance. |
| Incident intake | Can the agent collect the required details and create the expected record? | The agent returns an incident number, and the record represents the employee and reported issue correctly. |
| Assignment | Do the incident attributes match the intended active assignment rule? | The record is assigned to the expected user or queue. |
| Routing | Can the pilot fulfiller receive the assigned work? | The fulfiller has the required access, queue membership, presence, availability, and capacity. |
| Resolution | Can the fulfiller apply and verify a known fix? | The record contains the verified outcome, a useful resolution summary, and the appropriate status. |
Use the pilot users' own accounts. An admin-only test doesn't prove that employee access, portal deployment, record permissions, assignment, and routing work together. For the canonical portal and employee user procedures, see Create and Publish an Employee Services Portal Site, Create Employee Users for the Employee Services Portal, and Set Up Permission Sets and Access for the Employee Services Portal.
Connect the End-to-End Journey
Treat the first incident as a chain of accountable components rather than one Agentforce action. A successful incident-creation action doesn't prove that assignment, routing, resolution, and employee updates work.
| Journey Stage | Responsible Component | Evidence to Confirm |
|---|---|---|
| Employee reports the issue | Deployed IT Service Employee agent and its configured incident action | The conversation collects missing details, creates an incident, and returns its number. |
| Incident receives operational context | Incident configuration, including any configured priority logic | Reporter, subject, description, impact, urgency, priority, and status support the pilot process. |
| Incident reaches the correct team | Assignment rule or another configured assignment method | The record owner is the intended user or queue. |
| Work reaches a fulfiller | Omnichannel Routing | The intended fulfiller is available to receive and accept the work item. |
| Fulfiller investigates and fixes the issue | Incident workspace, Knowledge, and optional IT Service Fulfiller agent assistance | The fulfiller understands the issue, uses an appropriate solution, and verifies the result. |
| Resolution becomes visible | Incident lifecycle and configured notifications or employee ticket access | The record contains the resolution, and the employee can retrieve the new status or receives an update. |
Agentforce can create the incident, but incident creation doesn't by itself assign the record to a queue or deliver the work item to a fulfiller. Assignment and Omnichannel Routing remain distinct controls. See How an Incident Record Moves Through Omnichannel Routing, Assignment Rules for IT Services, and Omnichannel Routing for IT Services.
Validate Incident Creation in the Employee Services Portal
Ask the pilot employee to use an explicit incident request for the selected issue. Confirm that the agent asks for missing information rather than creating an incomplete record, creates the incident only after the employee's intent is clear, and returns a ticket number.
Inspect the resulting record for the employee identity, reported method, subject, description, impact, urgency, priority, status, and owner. The exact fields depend on the configured process. The record requires enough information for assignment and resolution without the admin repairing the intake manually.
For the supported IT Service Employee agent incident actions and example interactions, see AI Actions for Employee Request Management. For agent creation, activation, and testing, see Create and Configure the IT Service Employee Agent. For portal deployment, see Additional Settings for IT Service Agents.
Validate Queue Assignment and Omnichannel Routing
First, confirm that the created incident reaches the intended user or queue. Then, only if the owner is a queue, confirm separately that Omnichannel Routing offers the work item to the pilot fulfiller. If the owner is a user, skip Omnichannel because the fulfiller already has the record. If either transition fails, identify the failing boundary instead of treating the complete journey as an Agentforce failure.
Use the incident attributes and active rule criteria to investigate assignment. For routing, check queue membership, permissions, presence status, availability, capacity, workload, skills when configured, and the routing setup. A correctly assigned record can still remain undistributed when no eligible fulfiller is available.
If automatic priority is part of the pilot, verify that the resulting priority reflects the configured impact and urgency logic. Keep this check separate from assignment because priority calculation doesn't determine whether routing is operational. See Configure Automatic Priority Assignment for IT Services for the canonical configuration procedure.
Validate Incident Resolution in Agentic IT Service Desk
The pilot fulfiller accepts the work and reviews the incident context and relevant articles in the Knowledge component before selecting and applying a fix. The fulfiller verifies the result against the original symptoms. The IT Service Fulfiller agent can assist with configured actions such as checking attributes, summarizing the incident, finding related records, or proposing a resolution.
The fulfiller remains accountable for the resolution. Generated summaries and proposals are supporting evidence, not proof that the issue is fixed. After verification, the fulfiller records what changed and how the outcome was confirmed. The fulfiller then uses Resolved when the issue is fixed or Closed when the incident lifecycle is complete under the company's process.
For the canonical incident lifecycle and resolution procedure, see Manage Incidents for IT Services. For optional agent assistance, see Create and Configure the IT Service Fulfiller Agent and AI Actions for Incident Management. For the role of Knowledge in employee and fulfiller experiences, see Work with Lightning Knowledge for IT Services.
Confirm the First-Incident Milestone
The first-incident journey passes only when all of these outcomes occur without admin intervention in the employee or fulfiller tasks:
- The pilot employee can access the Employee Services portal and create the incident.
- The incident accurately represents the employee's issue and returns a ticket number.
- The configured assignment method sends the incident to the intended user or queue.
- If the incident is assigned to a queue, Omnichannel Routing offers the work item to an eligible pilot fulfiller. If the incident is assigned directly to a user, the assigned fulfiller can access the record.
- The fulfiller accepts the work, applies an appropriate fix, and verifies the result.
- The incident records the resolution and appropriate lifecycle status.
- The employee can retrieve the updated ticket status or receives the configured update.
If the journey fails, record the last successful boundary and the first failed boundary. These boundaries help identify whether the failure occurred in employee access, agent behavior, incident data quality, assignment, routing, fulfiller access, resolution, or employee communication. For notification channels and behavior, see Notifications for IT Services.
Completing one traceable incident journey demonstrates first value and pilot readiness. Evidence of adoption comes from repeatable successful journeys across a broader group of employees and fulfillers over time.
Separate Incident Creation from Employee Self-Service
A Knowledge answer or successful service catalog flow is a different employee outcome from incident creation. Successful self-service doesn't necessarily create an incident, enter assignment, or require Omnichannel Routing. Test it in a separate conversation with a distinct request and evaluate the answer or request record that the configured path is expected to produce.
Don't assume that standard Knowledge grounding creates and resolves an incident. If the intended experience requires one automated path to create a record, complete underlying work, and update its status, that behavior requires the corresponding configured action or flow. See Employee Self-Serves an IT Request in Slack or Microsoft Teams for the separate employee self-service adoption journey.
