Knowledge Graph

Overview

The Knowledge Graph is the structured layer ContractRabbit builds from your documents. It connects documents, matters, parties, document types, clause classes, extracted attributes, custom fields, citations, and saved reports.

Open Knowledge Graph in the app to search across this graph, create reusable fields, and build reports from the same data the Agent uses.

Query

The query tab combines full-text search, clause taxonomy, structured attributes, parties, document types, and matters. Use it to find documents or clauses by meaning and metadata:

• "Indemnification clauses in acquisition agreements"
• "Documents with Delaware governing law and New York forum"
• "NDAs where non-solicit duration is longer than two years"

Query results include source document links and matching field values so you can move from a graph result back to the underlying contract.

For more detail, see Query.

Fields

Fields are reusable data definitions. They can be:

• Built-in fields from document metadata and extracted attributes
• Mapped fields that normalize one or more existing attribute paths
• LLM fields that extract a value using a custom prompt
• Computed fields that derive a value from other fields

Fields can apply at the document level, clause level, or both. Reusing fields keeps reports and dashboard widgets consistent because the same definition is used everywhere.

For more detail, see Fields.

Reports

The reports tab lets you turn Knowledge Graph searches and fields into saved tables. Choose row scope, filters, target documents, and report columns, then preview and export the result.

Use reports for diligence trackers, renewal reviews, obligation inventories, precedent studies, and any recurring analysis where the same question needs to be rerun against updated documents.

For more detail, see Reports.

Agent integration

The Agent can use Knowledge Graph data when it answers questions, generates tables, compares attributes, and retrieves source clauses. If you ask for a report in natural language, the Agent can create a table from the same fields, attributes, and clause classes available in the Knowledge Graph.

Relationship to Documents and Data

• Documents is where you upload, review, edit, and manage individual documents
• Data is where admins configure document types, attributes, entities, and standards
• Knowledge Graph is where extracted data becomes searchable fields, reusable reports, and corpus-level analysis