How to create search index for an attribute and query it which has nested dynamic fields in mongoDB atlas search?

I want to create an index on qualityConfirmations attribute and query it. As qualityConfirmations has dynamic fields at the root level for eg. 228, 234, and inside that I need to query for "qualityConfirmation.qualityConfirmationComment", "qualityConfirmation.inspectDate" and other fields.

  {
    "_id": "4fa10c9b-099b-4269-92c9-ed22095f80c5",
    "qualityConfirmations": {
      "228": {
        "qualityConfirmation": {
          "delayGroupCode": "05",
          "estimatedDelayDays": 42,
          "inspectDate": {
            "$date": "2022-04-26T00:00:00.000Z"
          },
          "qualityConfirmationCode": "R",
          "qualityConfirmationComment": "QUALITY CONFIRMATION ACTIVITIES",
          "qualityConfirmationDescription": "RELEASED",
          "qualityConfirmationYear": 2022,
          "repairDate": {
            "$date": "2022-05-13T00:00:00.000Z"
          },
          "stopSaleNumber": "228"
        }
      },
      "234": {
        "qualityConfirmation": {
          "delayGroupCode": "40",
          "estimatedDelayDays": 294,
          "inspectDate": {
            "$date": "2022-05-26T00:00:00.000Z"
          },
          "qualityConfirmationCode": "R",
          "qualityConfirmationComment": "QUALITY CONFIRMATION ACTIVITY",
          "qualityConfirmationDescription": "RELEASED",
          "qualityConfirmationYear": 2022,
          "repairDate": {
            "$date": "2022-09-23T00:00:00.000Z"
          },
          "stopSaleNumber": "234"
        }
      }
    }
  },
  {
    "_id": "4fa10c9b-099b-4269-92c9-ed22095f80c6",
    "qualityConfirmations": {
      "237": {
        "qualityConfirmation": {
          "delayGroupCode": "39",
          "estimatedDelayDays": 287,
          "inspectDate": {
            "$date": "2022-06-02T00:00:00.000Z"
          },
          "qualityConfirmationCode": "R",
          "qualityConfirmationComment": "QUALITY CONFIRMATION ACTIVITY",
          "qualityConfirmationDescription": "RELEASED",
          "qualityConfirmationYear": 2022,
          "repairDate": {
            "$date": "2022-11-07T00:00:00.000Z"
          },
          "stopSaleNumber": "237"
        }
      },
      "246": {
        "qualityConfirmation": {
          "delayGroupCode": "34",
          "estimatedDelayDays": 252,
          "inspectDate": {
            "$date": "2022-07-19T00:00:00.000Z"
          },
          "qualityConfirmationCode": "R",
          "qualityConfirmationComment": "QUALITY CONFIRMATION ACTIVITY",
          "qualityConfirmationDescription": "RELEASED",
          "qualityConfirmationYear": 2022,
          "repairDate": {
            "$date": "2022-09-23T00:00:00.000Z"
          },
          "stopSaleNumber": "246"
        }
      }
    }
  },
  {
    "_id": "4fa10c9b-099b-4269-92c9-ed22095f80c7",
    "qualityConfirmations": {
      "265": {
        "qualityConfirmation": {
          "delayGroupCode": "19",
          "estimatedDelayDays": 147,
          "inspectDate": {
            "$date": "2022-10-19T00:00:00.000Z"
          },
          "qualityConfirmationCode": "R",
          "qualityConfirmationComment": "ABC CONFIRMATION ACTIVITY",
          "qualityConfirmationDescription": "RELEASED",
          "qualityConfirmationYear": 2022,
          "repairDate": {
            "$date": "2022-11-08T00:00:00.000Z"
          },
          "stopSaleNumber": "265"
        }
      }
    }
  }
]

I've used below schema and query for same.

  "mappings": {
    "dynamic": true,
    "fields": {
      "qualityConfirmations": {
        "type": "document",
        "dynamic": true,
        "fields": {
          "*": {
            "dynamic": true,
            "type": "document",
            "fields": {
              "qualityConfirmation": {
                "type": "document",
                "dynamic": false,
                "fields": {
                  "qualityConfirmationComment": {
                    "type": "string",
                    "analyzer": "earmarkAnalyzer"
                  }
                }
              }
            }
          }
        }
      }
    }
  },
  "analyzers": [
    {
      "charFilters": [],
      "name": "earmarkAnalyzer",
      "tokenFilters": [],
      "tokenizer": {
        "type": "keyword"
      }
    }
  ]
}



**QUERY:**

[
  {
    $search: {
      index: "default",
      compound: {
        must: [
          {
            phrase: {
              path: {
                wildcard:
                  "qualityConfirmations.*.qualityConfirmation.qualityConfirmationComment"
              },
              query: ["ABC CONFIRMATION"],
            }

          },
        ],
        mustNot: [],
        should: [],
        minimumShouldMatch: 0,
      },
    },
  },

]

But the result is not coming as expected. For example, If I want to use exact match , it still brings up all the results matching input, I've also used KEYWORD analyzer to create single token but it is not working. I want to search for both string and date Types.

Please find below the atlas playground link for same .

https://search-playground.mongodb.com/tools/code-sandbox/snapshots/676030e03206573cca5920b8