Enrichment policy based on multiple fields / complex query?

[akb]

I am investigating a way to normalize / prettify geo data during ingestion.
I realize there is a way to normalize by lat / lon, however we don't always have these properties availble.

Master geo data:

{
  "country": "de",
  "region": "Bayern",
  "region_alternate": ["Bavaria"],
  "city": "München",
  "city_alternate": ["Munich", "Muenchen"]
}

Incoming data may look like:

{
  "country": "de",
  "region": "Bavaria",
  "city": "Munich"
}

What I need to achieve is to store this geo version of incoming data in index:

{
  "country": "de",
  "region": "Bayern",
  "city": "München"
}

To find this in master geo data , I can use the following query:

{
  "explain": false,
  "query": {
    "bool": {
      "filter": [
        {
          "terms": {
            "country_code": [
              "de"
            ]
          }
        }
      ],
      "must": [
        {
          "multi_match": {
            "fields": [
              "region",
              "region_alternate"
            ],
            "query": "Bavaria"
          }
        },
        {
          "multi_match": {
            "fields": [
              "city",
              "city_alternate"
            ],
            "query": "Munich"
          }
        }
      ]
    }
  }
}

Is it possible to create an enrichment policy that is based on a more complex lookup rather than just a term query on a single field?

Any other ideas on how this geo normalization / prettifying could be achieved during data ingestion?

Thanks for any help or pointers to look into!