Skip to main content
A knowledge graph powers every world. Worlds organizes information in a graph-based structure rather than rigid tables, allowing you to model complex relationships with precision.

​
Knowledge primitives

To build with Worlds, you must understand the three fundamental building blocks of knowledge.

​
Items

Items are the distinct “things” in your world—a person, a piece of code, or a company.
  • Identification: A unique IRI, or Internationalized Resource Identifier, identifies every item.
  • Classes: Categorize items by their type, such as User, Project, or Task.

​
Properties

Properties are the “verbs” or “connectors” that define how items relate.
  • Examples: worksOn, managerOf, hasPriority.
  • Ontology: The set of defined properties forms the “grammar” of your world, ensuring your agents use consistent language.

​
Facts

A fact occurs when you connect two items using a property. This follows the standard W3C RDF triple structure:
ComponentExample
SubjectUser:Ethan
PredicateisWorkingOn
ObjectProject:Worlds-API

​
Why graphs matter

Table vs Node-Edge Graph

​
Meaning at scale

Historically, there has been a trade-off between semantic interoperability and analytical performance.
  • Semantic precision: RDF (Resource Description Framework) uses standard URI-based triples to define canonical vocabularies and ensure global interoperability. It is the gold standard for representing knowledge, but traditional RDF databases can be slow for traversing dense connection paths.
  • Analytical scale: Continuous AI workflows (like GraphRAG) require high-velocity, graph-native operational infrastructure to traverse complex relationships quickly under pressure.
Worlds bridges this gap. It leverages the interoperable, atomic meaning of standard RDF triples while functioning underneath as the high-speed computational substrate required for agentic workflows, giving you both semantic precision and analytical scale. Knowledge graph statements represent facts. Unlike statistical LLM weights, Worlds retrieves specific, auditable relationships.
  • Malleability: Mutate and fork graphs in real-time.
  • Traceability: Every claim provides a symbolic path back to its source.
  • Evolution: New facts can surgically update or override old ones to maintain grounded truth.

​
Technical context

Worlds uses the RDF 1.1 standard for universal portability. You can interface with the graph through the Worlds SDK, which abstracts away the complexity of raw SPARQL queries.