Skip to main content
Instances are the individual data records (nodes and edges) stored in CDF’s knowledge graph. Nodes represent entities (such as equipment, locations, or work orders) while edges represent relationships between them. Instances can be filtered, searched, aggregated, queried, and synced. For the conceptual overview, see Spaces and instances and Query features.

​
What are instances

Instances are the concrete data that populates your containers and views. Each instance conforms to the schema of its container. When you write data through a view, you create or update instances. When you query a view, you read instances.

​
Nodes and edges

Nodes are instances that represent entities. Edges are instances that connect two nodes and represent relationships. Together, nodes and edges form a graph structure that you can traverse and analyze. Both are stored in containers and exposed through views.
Edges reference source and target nodes by their instance IDs. You can query edges by property, by node reference, or by graph traversal patterns.

​
Querying instances

You can query instances using filters, search, and aggregation. Filter expressions allow you to narrow results by property values. Search supports full-text and semantic queries. Aggregation lets you compute counts, sums, and other statistics over instances.

​
Key capabilities

  • Create and update — Insert or upsert nodes and edges through views.
  • Delete — Remove instances when they are no longer needed.
  • List and filter — List instances with optional filters on properties.
  • Search — Search instances by text or semantic similarity.
  • Aggregate — Compute aggregations over instances.
  • Sync — Sync instances from external systems in bulk.
Last modified on April 23, 2026