Data modeling with the industrial knowledge graph

Who is this for?

All new customers and existing customers building next-generation solutions using CDF’s industrial knowledge graph. In data modeling, assets are represented as instances using the CogniteAsset concept in the core data model—with support for hierarchies, relations, and flexible schemas. To learn more about data modeling for the knowledge graph, see the Data modeling section.

Choose your path

I'm building applications

Use REST and GraphQL APIs to query and mutate data. Start with the GraphQL schema service and Query features.

I'm designing data pipelines

Define schemas, ingest data, and build high-volume pipelines. Start with Containers and views, Ingestion, and Records and streams.

Key APIs

Spaces — Define data boundaries. See Spaces and instances.
Instances — Nodes and edges. See Query features.
Containers — Schema definitions. See Containers and views.
Views — Queryable projections. See Containers and views.
Data models — Grouped views. See Data models.

For high-volume bulk data: Records and Streams. See About records and streams.

Reference

Examples and best practices — Building asset hierarchies, extending the core model, integrations, performance, and CI/CD.
Limits and restrictions — Resource limits, property value limits, and concurrency for data modeling.
Core data model — CogniteCore building blocks (Asset, TimeSeries, File, and more).
Process industries data model — CogniteProcessIndustries extension for maintenance orders, operations, and notifications.

Data-plane APIs

When working with industrial data through data modeling, you also use these APIs for the underlying data operations:

Time series and datapoints

Insert, retrieve, and query time series datapoints.

Datapoints

Insert, retrieve, and delete data points in time series.

File content

Upload and download file content.

Last modified on August 28, 2026

⌘I

​Who is this for?

​Choose your path