Build APIs as part of the system.
Create and expose APIs directly from the business logic that powers your AI System. APIs are not a separate layer around the application; they are part of the same runtime.

From logic to API
Define business logic once and expose it through an API without creating a separate service around it.
The API and the logic it exposes share the same runtime, data, dependencies, and governance.
Build without the deployment cycle
Create or change an API without rebuilding and redeploying the application.
Changes are interpreted by the AVAP runtime and become available as the system evolves.
APIs for any consumer
APIs can coexist with MCP and A2A interfaces within the same AI System.
Expose the same capabilities to:
- Applications and services
- External systems
- AI applications
- AI agents
- MCP tools
Contracts and governance
Define API contracts and control how capabilities are exposed and consumed.
Manage access, permissions, and policies through the same governance model used across the system.
APIs as system capabilities
An API is an interface to a capability, not a separate application component.
The underlying business logic remains connected to the rest of the AI System and can be reused through other interfaces.