Metadata-Driven Systems
Metadata describes much of a system’s structure and behavior. In 3D, it covers more than tables and columns.
It can define relationships, validation, history, auditing, forms, commands, navigation, security, and data exchange.
The same definitions are used to generate database structures, constraints, interfaces, forms, navigation, audit mechanisms, and other application components.
Metadata at the Core
Generation in the 3D Platform
Requirements → Concepts → Shared Structures → Local Structures → Validated Metadata → Generation
-
Start with the business unit’s requirements and identify its business concepts.
-
Reuse suitable shared structures and add the structures the business unit needs.
-
Use AI to help interpret requirements.
-
Use deterministic processing for results that can be derived mechanically.
-
Validate the metadata before generation.
Why it matters
One set of definitions can drive the structures and supported behavior across the system. We have implemented and exercised system generation in the 3D Platform, the software used to demonstrate the architecture.
What has been demonstrated
The implemented pipeline starts with requirements and generates systems. Reverse engineering existing systems is separate work. Other proposed automation remains planned until it has been implemented and demonstrated. Generated environments show that the implementation works in those environments; they do not establish production readiness at organizational scale.