Using agentic workflows in Now Assist for CMDB
Summarize
Summary of Using agentic workflows in Now Assist for CMDB
Now Assist for CMDB offers agentic workflows that enable users with thesncmdbuserrole to interact with the Configuration Management Database (CMDB) more efficiently through natural language requests and AI-driven processes. These workflows streamline tasks such as creating configuration items (CIs), improving CMDB governance, and searching for CIs by attributes and relationships. AI agents support these workflows by automating data validation, certification, ownership assessment, and lifecycle management, enhancing the accuracy and trustworthiness of CMDB data.
Show less
Agentic Workflows and Their Functions
- Create configuration item: Allows manual creation of a CI using natural language input. It verifies the CI class, checks Identification and Reconciliation Engine (IRE) rules for required attributes, prevents duplicates, and creates the CI accordingly.
- Provide advice on CMDB governance: Supports data administrators and owners in improving CMDB data accuracy, completeness, and health by methodically guiding governance efforts.
- Search CMDB: Enables searching for CIs by various attributes using natural language. It determines the appropriate search type (keyword, single-table, or multi-table with relationship navigation) and infers CI relationships to refine queries.
Key AI Agents Supporting Workflows
- CMDB CI creator: Creates new CI records following workflow validations.
- CMDB data certification and attestation manager: Retrieves and applies data certification and attestation policies for CI classes.
- CMDB data ownership manager: Assesses and validates CI ownership data, including user and group references, and classifies ownership quality.
- CMDB health metrics manager: Provides CMDB health dashboard insights and guides setup if health monitoring jobs are unconfigured.
- CMDB life cycle manager: Manages lifecycle policies for CI classes, including retrieval and creation of lifecycle and retirement policies.
- CMDB principal class manager: Helps designate principal classes to optimize visibility and management of commonly used CIs.
- CMDB search AI agent: Performs keyword or table queries to locate specific CIs and presents search results.
- CI summarizer AI agent: Generates concise summaries of key CI data, including discovery, ownership, related incidents, changes, and vulnerabilities.
Practical Considerations for ServiceNow Customers
- To run AI agents autonomously, duplicate and activate the agentic workflow, activate all included AI agents, and enable triggers if automatic invocation is desired.
- Security is enforced via Access Control Lists (ACLs) that govern AI agent and workflow execution, supporting Run As capabilities for dynamic or AI user identities.
- Customers can configure which AI model providers (e.g., Now LLM Service, Azure OpenAI) are available and set preferences at the skill level within the Now Assist Admin console.
- Agentic workflows simplify complex CMDB tasks, reducing manual effort and improving data quality, which is crucial for reliable IT service management and operational decision-making.
Getting Started
ServiceNow customers can leverage these agentic workflows and AI agents to:
- Create validated CIs efficiently using natural language input.
- Obtain structured advice and governance guidance to maintain CMDB integrity.
- Search the CMDB effectively by attributes and inferred CI relationships.
These capabilities help improve CMDB data trustworthiness and usability, enabling better governance, lifecycle management, and operational insights.
Users with the sn_cmdb_user role can access several Now Assist for CMDB agentic workflows.
Agentic workflows
| Agentic workflow name | Description | Available AI agents |
|---|---|---|
| Create configuration item | Occasionally, you might create a CI manually. To help you, the Create configuration item agentic workflow accepts your natural language request and verifies that it understands which class the new CI should belong to. The workflow then checks Identification and Reconciliation engine (IRE) rules to determine the required attributes for the CI and requests that information. After you provide sufficient data, the workflow ensures that the proposed CI includes the attributes that you requested, complies with IRE rules, and is not a duplicate. The workflow then creates the CI. For more information, see Create a CI using Now Assist | CMDB CI creator |
| Provide advice on CMDB governance | Data governance can be an overwhelming task. The Provide advice on CMDB governance agentic workflow supports data admins and owners by methodically working through the many-faceted process of improving CMDB data accuracy, completeness, and health. The objective is to promote that users can trust the data that they use for their work. For more information, see Getting advice from Now Assist on CMDB governance. |
|
| Search CMDB | The Search CMDB agentic workflow enables you to search for CIs by specifying any of several attributes of the CI of interest. The workflow accepts your natural language request, verifies your search goal, and then, depending on the information you provided, generates a keyword search, a single-table search with dot walks, or a multi-table search that involves relationship navigation. The workflow can infer CI relationship data to generate an appropriate query. For more information, see Use Now Assist to search the CMDB for CIs. | CMDB CI search |
AI agents used by Now Assist for CMDB
- There might be AI agents installed with the Now Assist application that are not used in agentic workflows. To learn how to see all agents that are available on your instance, see Find AI agents.
- To find agents that might not be installed on your instance, visit the AI Agent Marketplace on the ServiceNow Store.
- CMDB CI creator AI agent
- The CI creator AI agent creates a new configuration item (CI) record in the CMDB.
- CMDB data certification and attestation manager AI agent
- The CMDB data certification and attestation manager AI agent retrieves the data certification and attestation policies for a CI class and for similar classes. It can also create a data certification and an attestation policy. For more information, see Data Certification.
- CMDB data manager AI agent
- The CMDB life cycle manager AI agent helps admins to promote that CMDB life-cycle policies are in place for principal classes or for specified classes. For more information, see Working with CMDB Data Manager.
- CMDB data certification and attestation manager AI agent
-
The CMDB data certification and attestation manager AI agent retrieves and applies data certification and attestation policies to manage the data certification and attestation of CI classes.
- CMDB data ownership manager AI agent
-
The CMDB data ownership manager AI agent helps you to assess the quality of CI ownership data for a CI class to determine ownership, validate ownership, and evaluate ownership quality. You can validate fields that reference sys_user or sys_user_group like Assigned To, Managed by Group (default), Support Group, and so on. The agent answers the following questions:
- Is CI ownership specified? Is the Managed by group attribute value set for the CI? Does it indicate an active and correctly configured record? The value of the system property for ownership is a default field for the evaluation. For the Managed by group CI attribute, you can specify the percentage values that classify the result as Poor, Average, or Healthy. For more information, see Property settings for Now Assist for CMDB.
- If you provide a user referencing value: Does the User Referencing value indicate an active record? This evaluation identifies CIs that have inactive users and CIs with no User value.
- If you provide a user referencing value: Does the Group Referencing value indicate an active record? This evaluation identifies CIs that have inactive groups, empty groups (no users), and CIs with no Group value.
- CMDB health metrics manager AI agent
- The CMDB health metrics manager AI agent displays information on CMDB Health Dashboard jobs, health metrics status, and health scores. If the jobs are not configured, the workflow provides a link to the setup instructions. For more information, see CMDB Health.
- CMDB life cycle manager AI agent
- The CMDB life cycle manager AI agent performs actions related to life-cycle management of CI classes: retrieve the life-cycle management policy and retirement definition for a CI class and create a life-cycle policy.
- CMDB principal class manager AI agent
- The CMDB principal class manager AI agent determines whether you can set Cl classes as principal, suggests principal classes, and can set or unset a class as principal. Setting a class as a principal class is a practical way to save time because you can configure ServiceNow AI Platform products to display only principal classes in list views. The workflow helps you to decide which CIs you're most likely to work with day-to-day and then provides a link to enable you to set the associated classes as principal classes. For more information, see Update class list in the Principal Class filter.
- CMDB search AI agent
- The CMDB search AI agent searches the CMDB for a particular CI using single table query search or keyword query search and then displays a link to search results.
- CI summarizer AI agent
- View a concise summary of key CI data. You can select the CI on a CI form, in a workspace page, or on any list view. The summary can include discovery data, ownership, and key related items such as open incidents, alerts, problems, upcoming change requests, and security vulnerabilities. Additionally, the summary lists the service instances that the CI is part of.
You can use Now LLM Service, Now LLM Long Term Stable models (LTS), Azure OpenAI, Google Gemini or Anthropic Claude on AWS as the AI model provider for all Now Assist skills and AI agents. Use the Configuration Controls in AI Control Tower to define which options are available, then set the skill-level preferences in the Now Assist Admin console. For more information, see Large language models on the ServiceNow AI Platform®.
Enable security implementation to execute AI agents and agentic workflows through Access Control Lists (ACLs) and user identities. ACLs provide the Run As capability to let agents and agentic workflows execute actions either as a dynamic user or as an AI user. For more information, see Implement access control in Now Assist AI agents
Running AI agents autonomously
To run the AI agents autonomously, you must first duplicate the agentic workflow and then proceed with the following steps:
- Activate the agentic workflow.
- Activate all agents within the agentic workflow.
- Activate the trigger to invoke the agentic workflow automatically. If you prefer to invoke it manually, activating the trigger isn’t necessary.