...

Story 2: Multiple Sources

Option 1: Introduce different types of connections. One for data flow, one for control flow

...

You could also say that connections into a source imply control flow, or connections into an action imply control flow. 

Story 2: Spark ML in a pipeline

Add a plugin type "sparkML" that is treated like a transform.  But instead of being a stage inside a mapper, it is a program in a workflow.  The application will create a transient dataset to act as the input into the program, or an explicit source can be given.

{
  "stages": [
    {
      "name": "customersTable",
      "plugin": {
        "name": "Database",
        "type": "batchsource", ...
      }
    },    
    {
      "name": "categorizer",
      "plugin": {
        "name": "SVM",
        "type": "sparkML", ...
      }
    },
    {
      "name": "models",
      "plugin": {
        "name": "Table",
        "type": "batchsink", ...
      }
    },
  ],
  "connections": [
    { "from": "customersTable", "to": "categorizer" },
    { "from": "categorizer", "to": "models" }
  ]
}