API / Sealarca service
Receives the request and prepares its processing within the service.
INFRASTRUCTURE
Understand how Sealarca works technically.
This page describes the path of a request through the Sealarca service: API and access control, orchestration and allocation, Vault execution environment, model and operational measurement.
Protection follows your request to the response.
- Sealarca service
- Vault environment
- Selected model
- Vault proof
SEALARCA SERVICE
The path of a request
Each step describes the public role of the layer that receives, authorizes, prepares and executes the request.
USER ENVIRONMENT
SEALARCA SERVICE
- 1
Client / application
Direct API, Desk, Bridge or software integrating the API.
- 2
API / Sealarca service
Receives the request and prepares its processing within the service.
- 3
Access control
Checks the conditions required to authorize the request.
- 4
Orchestration
Determines a route compatible with the request and operational configuration.
- 5
Allocation
Allocates the resources required to run the selected route.
- 6
Vault execution environment
The relevant route uses the protected Vault environment to process the necessary content.
- 7
Model
Processes the necessary content according to the route and produces the response.
- 8
Response
The response returns to the application through the Sealarca service.
A request starts in the client environment, reaches the Sealarca service, passes through access control and then orchestration and allocation, is processed by the relevant route in the Vault execution environment with the selected model, and the response then returns to the client.
The role of each layer
These roles describe the service at a public level; sensitive operational details remain outside this page’s scope.
Access control
Checks the conditions required to authorize the request.
Orchestration and allocation
Determines compatible execution from the request and operational configuration.
Vault execution environment
Provides the protected environment used by relevant routes.
Model
Processes the necessary content according to the route and produces the response.
Content and operational data
The service distinguishes content needed for processing from operational information needed to run the service.
Content
- Prompts and instructions
- Documents or content sent
- Generated response
Operational data
- Requested or used model
- Token volume
- Request status
- Execution duration
- Calculated cost
Operational information supports service operation, usage measurement and billing. This page does not set a retention period.
Scope of this page
What this page describes
- Request reception and access control
- Orchestration and allocation
- Vault execution environment and model
- Response return and operational measurement
What this page does not replace
- Safeguards and limits → Security
- Legal commitments → legal pages
- Versioned Desk and Bridge behavior → Application data
- Developer integration → API documentation
Dynamic catalog and routes
Available models, published characteristics and associated routes are fed by the service’s operational configuration. This page explains the mechanism without duplicating the catalog.
View models and pricing- 1
Available catalog
Models and published characteristics come from the operational configuration.
- 2
Service configuration
The configuration defines the usable routes.
- 3
Compatible route
The service associates the request with the corresponding route.
- 4
Execution
The selected route runs and returns the response.
Ready to use Sealarca?
Create your account, then choose Desk, Bridge or the API according to how you work.