Skip to main content

Overview

Configure and deploy out-of-the-box AI Agents and Teams that work directly within ServiceOps to answer questions, handle requests, and support technicians — without any custom development.

AI Studio provides administrators with a centralized platform to configure and manage AI Agents and AI Teams. ServiceOps provides a fixed set of out-of-the-box (OOB) AI Agents and Teams. Creating new agents or teams and cloning existing ones is not supported. Administrators can attach Knowledge Collections, enable PII detection, and preview each agent or team before it goes live.

What Are AI Agents and AI Teams?

An AI Agent is an individual, autonomous AI entity pre-built to perform specific tasks within ServiceOps. Each agent operates with its own built-in persona, operating principles, knowledge sources, and tools. Agents interact with users and technicians to handle queries, process requests, and execute actions without manual intervention. They are the foundational units of agentic AI in ServiceOps.

AI Teams are structured groups of AI Agents orchestrated to achieve complex, multi-step goals that go beyond what a single agent can accomplish. A designated manager agent coordinates the team, assigning subtasks, routing requests to the most capable agent, and consolidating outcomes. Teams enable collaborative AI workflows where diverse agent skill sets work together to resolve sophisticated service scenarios.

Agent vs. Team: Which to Use?

Use the table below to decide which construct fits your scenario. When in doubt, start with a single agent to validate your use case and move to a team once you identify multi-domain or multi-step patterns.

ScenarioRecommended Approach
Single, well-defined task such as answering FAQs or checking ticket statusSingle AI Agent
Task requires a consistent, specialized persona such as an HR policy assistantSingle AI Agent
Pilot or early-stage AI deployment with limited scopeSingle AI Agent
Request spans multiple domains such as onboarding involving IT, HR, and facilitiesAI Team
Workflow requires escalation or handoff between agentsAI Team
Resolution depends on sequential subtasks handled by different specialistsAI Team
FeatureAI AgentAI Team
RolePre-built autonomous agent for a single service functionPre-built group of specialized agents coordinated by a manager agent
TasksSpecific, well-defined tasksComplex goals requiring collaboration across multiple domains
OperationOperates independentlyCoordinated by a manager agent
ConfigurationKnowledge Collections and PII detectionTeam-level Knowledge Collections and PII detection
Best forFocused automation and specialized supportMulti-step workflows and cross-domain service requests

Key Capabilities

  • Knowledge Integration: Attach Knowledge Collections so agents and teams can reference trusted, organization-specific content when responding to users.
  • Tool Visibility: View the read-only list of tools available for each agent or team, such as modules they can interact with, for example, Request and Service Catalog.
  • Responsible AI Policies: Apply PII detection at the agent or team level to ensure sensitive data is masked or blocked during interactions.
  • Version History: Every configuration change is saved with a timestamp. Access the version history from the agent or team configuration page to review past changes and restore a previous version if needed.

What Can Be Configured?

Each OOB AI Agent and Team comes with a fixed persona, built-in tools, and operating principles. The following table clarifies what administrators can configure and what is read-only:

SettingConfigurable?
Knowledge Collections attached to the agent or teamYes
PII Detection (Mask or Block, Input or Output)Yes
Agent or team persona and instructionsNo — read-only
Tools available to the agent or teamNo — read-only
Agent or team name and typeNo — read-only

Who Uses AI Studio?

PersonaRole in AI Studio
IT AdministratorConfigures OOB agents and teams by attaching Knowledge Collections and enabling PII detection; monitors usage via Governance
IT TechnicianInteracts with deployed agents through the Technician Portal via Ask AI to get AI-assisted resolution support
End UserEngages with AI agents via the Support Portal for self-service request handling, status updates, and guided troubleshooting

Prerequisites

Before using AI Studio, ensure the following are in place:

  • AI license is active: AI Studio features require a valid ServiceOps AI license.
  • Roles and permissions configured: The admin must have Manage AI Studio permissions assigned under Admin > User > Roles > Admin Module.
  • Governance settings reviewed: Organization-level AI settings including tools access and responsible AI policies should be configured before deploying agents. See Governance: General Settings.
  • Knowledge Collections ready (optional but recommended): Agents and Teams perform better when backed by curated knowledge sources. See Knowledge Collections.

How It Works

AI Studio follows a straightforward setup-to-deployment flow:

  1. Configure Governance: Set organization-level AI policies, tools access, and responsible AI rules in the Governance Center before configuring agents. See General Settings.
  2. Configure an Agent: Select an OOB agent, attach the relevant Knowledge Collections, and enable PII detection as needed.
  3. Configure a Team (optional): When your use case requires collaboration across multiple domains, select an OOB team, attach Knowledge Collections, and enable PII detection.
  4. Preview: Use the Preview option on the agent or team configuration page to test how it responds to sample queries before it goes live. Check that responses are accurate, on-topic, and sourced from the correct Knowledge Collection.
  5. User Access: Once configured, agents and teams are accessible to users through the Ask AI button in the top navigation bar of the Technician and Support Portals, depending on the portal selection configured in General Settings.
  6. Monitor and Optimize: Use the Governance Analytics dashboard to track credit consumption, session quality, and agent performance over time.
Getting Started

Configure and validate one agent before setting up a team. This helps you understand how Knowledge Collections affect response quality and gives you a baseline to compare against when expanding to multi-agent teams.