Job System

Background jobs with real-time progress streaming are executed by defining job types and registering worker handlers that process tasks asynchronously. Progress updates are pushed to clients via Server-Sent Events (SSE), enabling live monitoring of long-running operations.

Long-running operations like large data imports and complex report generation are offloaded from the main application thread using asynchronous task execution. The job system manages the lifecycle, allowing the main thread to remain unblocked while workers process heavy computations.

Asynchronous task states are tracked through a built-in state management system that monitors jobs across stages like idle, running, stopped, succeeded, and failed. This allows you to query the current status and handle state transitions for distributed task management.

Custom workers for specific background job types are defined and registered within the job system to handle targeted task processing. You create a new job, specify its type, and start it with an input payload that the designated worker handler consumes.

Immediate user feedback during AI model inference is provided through real-time event streaming via Server-Sent Events (SSE). By offloading the inference to a background job, users receive live progress updates instead of waiting for the entire computation to finish.

External dependencies are not required to manage distributed task processing, as the job system operates independently. It provides built-in worker registration, job state management, and streaming support without relying on external libraries or packages.