Data Models

Data Models are structured representations of the entities in a system, their attributes, and the relationships between them — the concrete artifact that says what information exists and how it is shaped. Where relational models trace back to E. F. Codd’s work in 1970, today the term covers everything from database schemas to the request and response objects an API exposes, all serving the same purpose: a shared, enforceable definition of the data.

• Entities, attributes, relationships - The three primitives that describe what a system knows and how those things connect.
• Schema as contract - A data model expressed as a schema (JSON Schema, XSD, Avro, or a database DDL) becomes a validation and versioning contract.
• Reusability across surfaces - The same model can back a database table, an event payload, and an API resource, keeping meaning consistent.
• Multiple paradigms - Relational, document, key-value, graph, and event models each shape data for different access patterns.

In API operations, data models are what your OpenAPI components/schemas, your event payloads, and your database tables all point back to, which is why reusing them rather than reinventing them per endpoint keeps an API coherent. A well-governed catalog of shared models reduces drift, makes changes reviewable, and — in the agentic turn — gives automated clients a stable, machine-readable structure to bind to. They are the output of good data modeling and the meaning behind whatever data format ultimately carries them.