Exploring AI Agent Topology Mapping
Learn how AI Agent Topology Mapping discovers AI infrastructure components across cloud platforms using patterns.
AI Agent Topology Mapping overview
AI Agent Topology Mapping extends the pattern-based discovery framework to identify and track AI-specific components in your environment. The application uses patterns to discover AI components from cloud platforms, populating the CMDB with configuration items (CIs). This approach provides centralized visibility into your AI infrastructure alongside traditional IT assets. For more information about how patterns work, see Discovery patterns used by ITOM Visibility.
- AI Agents: Intelligent entities that perform tasks and orchestrate AI workflows
- AI Models: Foundational models that power AI capabilities, such as large language models
- AI Prompts: Instructions and configurations that guide agent behavior and responses
- AI Tools: Functions and capabilities that agents can invoke to complete tasks, such as API integrations or data retrieval functions
AI Agent Topology Mapping users
The following user roles have access to patterns or pattern-related modules and can perform various actions. Note that customizing patterns requires basic knowledge of programming.
| User | Description |
|---|---|
| Discovery admin | Can view, create, edit, and publish patterns. The role enables users to run discovery, migrate probes or CAPI to patterns, and access discovery logs and dashboards. |
| PD user | Has read-only access to Discovery Pattern Log. |
| PD admin | Can view, create, edit, and publish patterns. |
| PDE viewer | Starting with Pattern Designer Enhancements version 3.9.0, users can view Command Validation Tasks, Command Validation Tasks Results, and Command
List. The pde_viewer can view the Command Validation Tool modules and related tables, but doesn't have permissions to modify or edit them. The pde_viewer role can
view the following tables only:
|
| PD MID | Not assigned to a user directly but to the MID Server record or the user under which the MID Server runs. The role enables the MID Server to interpret and run pattern-based probes. |
| MID Server | Can grant the MID Server access to the instance. |
AI Agent Topology Mapping workflow
The following workflow describes how a discovery administrator uses AI Agent Topology Mapping to discover and track AI infrastructure components.
- Install AI Agent Topology Mapping patterns from the ServiceNow Store.
- Configure cloud credentials with appropriate permissions for AI platforms.
- Create or update discovery schedules for environments containing AI resources.
- Run discovery to identify AI components and populate the CMDB.
- View discovered AI agents, models, prompts, and tools in CMDB tables.
- Monitor discovery logs to verify successful pattern execution.
- Schedule recurring discovery to maintain up-to-date AI infrastructure inventory.
AI Agent Topology Mapping benefits
| Benefit | Description |
|---|---|
| Centralized AI infrastructure visibility | Visibility into AI agents, models, prompts, and tools alongside other IT assets in the CMDB. |
| Automated discovery | Patterns run automatically during scheduled discovery without requiring separate configuration or manual inventory processes. |
| Multi-cloud support | Discovery of AI resources across multiple cloud platforms using a single application. |
| Security and compliance | Tracking of AI component versions and configurations to help maintain governance and compliance requirements for AI deployments. |
| Vulnerability management | Tracking of AI component versions and dependencies to identify security risks. |
| Change impact analysis | Understanding of relationships between AI components before modifying configurations. |