MuleSoft - Dataweave 2.0 - Split payload into seperate arrays based on type

Having some trouble splitting a payload up. I basically need to split the payload every time we come across the {"type": "AA"} object.

This is my input payload -

[
    {
        "type": "AA",
        "content": "someContent"
    },
    {
        "type": "AB",
        "content": "someContent"
    },
    {
        "type": "AC",
        "content": "someContent"
    },
    {
        "type": "MH",
        "content": "someContent"
    },
    {
        "type": "AA",
        "content": "someContent"
    },
    {
        "type": "AB",
        "content": "someContent"
    },
    {   
        "type": "AC",
        "content": "someContent"
    },
    {
        "type": "QP",
        "content": "someContent"
    }
]

And what I need to happen is every time we come across the {"type": "AA"} we need to create an array and include all proceeding objects in that array. So this is what the output would look like

[
    [{
        "type": "AA",
        "content": "someContent"
    },
    {
        "type": "AB",
        "content": "someContent"
    },
    {
        "type": "AC",
        "content": "someContent"
    },
    {
        "type": "MH",
        "content": "someContent"
    }],
    [{
        "type": "AA",
        "content": "someContent"
    },
    {
        "type": "AB",
        "content": "someContent"
    },
    {   
        "type": "AC",
        "content": "someContent"
    },
    {
        "type": "QP",
        "content": "someContent"
    }]
]

You can see from the output we have multiple arrays with the first element being the object {"type": "AA"}.. Any help in achieving this output would be greatly appreciated

Assuming that the input always start with type=="AA" we can reuse the clusterWhile() function from this great previous answer: How to Group consecutive dates together which are separated by hour in Dataweave? to group elements of an array while a condition is true.

%dw 2.0 
output json
fun clusterWhile<T>(elements: Array<T>,  criteria: (source:T, target:T) -> Boolean) = do {
    fun clusterLoop(elements, value, carrier, criteria) = 
        elements  match {
            case [] ->  carrier
            case [x ~ xs] ->  
                if(criteria(value, x)) do {
                    var updatedCarrier = carrier update {
                        case [-1] -> $ << x    
                    }
                    ---
                    clusterLoop(xs, x, updatedCarrier, criteria)
                }
            else 
                clusterLoop(xs, x, carrier << [x], criteria)
    }

    ---
    elements match {
        case [] -> []
        case [x ~ xs] -> clusterLoop(xs, x, [[x]], criteria)
    }
}    
---
payload 
    clusterWhile ((source, target) -> !(target."type" == "AA"))

Output:

[
  [
    {
      "type": "AA",
      "content": "someContent"
    },
    {
      "type": "AB",
      "content": "someContent"
    },
    {
      "type": "AC",
      "content": "someContent"
    },
    {
      "type": "MH",
      "content": "someContent"
    }
  ],
  [
    {
      "type": "AA",
      "content": "someContent"
    },
    {
      "type": "AB",
      "content": "someContent"
    },
    {
      "type": "AC",
      "content": "someContent"
    },
    {
      "type": "QP",
      "content": "someContent"
    }
  ]
]