Async model processing

**Describe the solution you'd like**
Queue all models, that should be processed. 
A sheduled task will optimize all pending models.

Add table to store current model status and path to result files if exists.

**Additional context**
Add any other context or screenshots about the feature request here.
