Data Modeling

Data Modeling is the practice of defining the entities, attributes, relationships, and constraints that make up the information in a system, so that data can be organized, stored, and accessed consistently. It typically moves through three levels of abstraction — conceptual (what things exist and how they relate), logical (the detailed structure independent of technology), and physical (how it actually lands in a database or interface).

• Three levels of abstraction - Conceptual, logical, and physical models let teams reason from business meaning down to storage.
• Entities and relationships - Entity-relationship (ER) modeling captures the things a system tracks and how they connect, the classic backbone of the discipline.
• Constraints and integrity - Keys, cardinality, and validation rules encode what valid data looks like before a single row is written.
• A design conversation, not a diagram - The real output is a shared understanding of the domain that outlives any one database or API.

In API practice, data modeling is the upstream work that decides what your resources, request bodies, and response schemas will look like long before you write an OpenAPI definition. Getting the data models right keeps an API’s surface coherent as it grows, makes versioning less painful, and gives downstream consumers — increasingly automated agents — a predictable structure to reason about. It pairs with data flow diagrams, which show how that modeled information moves, and with the data format that ultimately serializes it on the wire.