NOW Knowledge Graph on Moveworks
What is the ServiceNow Knowledge Graph?
The ServiceNow Knowledge Graph is a semantic data layer that maps enterprise tables, records, and the relationships between them. This connected model helps AI tools and virtual agents understand business context—not just retrieve isolated records—so they can answer questions that span multiple parts of the ServiceNow data model.
Learn more in the ServiceNow Knowledge Graph documentation.
Availability
This capability depends on MCP Workspace, which will provide the Moveworks-side experience for configuring and managing MCP connections and tools.
How will it work with Moveworks?
This integration will use the Model Context Protocol (MCP):
- An administrator sets up an MCP server in ServiceNow.
- The administrator exposes the appropriate ServiceNow Knowledge Graph capabilities as MCP tools.
- The Moveworks Assistant connects to that MCP server and invokes the tools when a user's request requires connected information from the ServiceNow instance.
- The Assistant uses the returned, permission-aware context to provide a helpful response in the user's conversational workflow.
ServiceNow provides a guided process for creating an MCP tool from a Knowledge Graph.
Example questions
The questions available to users will depend on the Knowledge Graph schemas and tools configured in each ServiceNow instance. Examples could include:
- “What business services depend on this application?”
- “Which critical incidents affected services owned by my team this quarter?”
- “What infrastructure supports the employee onboarding service?”
- “Which change requests are related to the configuration items involved in this incident?”
- “Who owns the service affected by this outage, and what systems depend on it?”
- “Show me the applications, business capabilities, and owners connected to this service.”
Instead of requiring users to navigate tables and manually piece together relationships, the Moveworks Assistant can invoke the configured Knowledge Graph tools and present the relevant connected information conversationally.