Workflow Deployment Architecture - Independent vs. Monolithic Deployment

[BigBoiAlex]

I'm evaluating Elsa for a project and have questions about the deployment model

I found this discussion about scalability which covers distributed runtime and scaling strategies, but I'm looking for clarification on workflow deployment patterns specifically.

Current Understanding:
From the documentation, it appears that Elsa follows a monolithic deployment approach where:

• All workflows are bundled with the application and deployed as a single unit
• The distributed features (DistributedWorkflowRuntime, DistributedLocking, DistributedCaching) are for horizontal scaling across multiple nodes of the same application instance

Questions:

1. Is this understanding correct, or does Elsa support deploying individual workflow definitions separately from the main application?
2. When I need to add a new workflow or update an existing one, does this require redeploying the entire application? Or is there a way to hot-deploy workflow definitions at runtime?
3. Can multiple versions of the same workflow definition coexist in a single Elsa deployment, with the ability to control which version new workflow instances use?

Use Case:
I'm looking for the ability to deploy workflow versions independently (similar to microservices deployment patterns) to avoid full application redeployments when only workflow logic changes.