The collection schema is something like:

{
  "name": "travelers",
  "fields": [
    {
      "name": "_id",
      "type": "string",
      "index": true,
      "optional": false
    },
    {
      "name": "first_name",
      "type": "string",
      "index": true,
      "optional": false
    },
    {
      "name": "last_name",
      "type": "string",
      "index": true,
      "optional": false
    },
    {
      "name": "email_address",
      "type": "string",
      "index": true,
      "optional": false
    },
    {
      "name": "employee_id",
      "type": "string",
      "index": true,
      "optional": true
    },
    {
      "name": "company_id",
      "type": "string",
      "index": true,
      "optional": true
    },
    {
      "name": "company_group_id",
      "type": "string",
      "index": true,
      "optional": true
    }
  ],
  "num_documents": 3685055,
  "token_separators": [
    "-"
  ]
}

An example of a typical request is:

GET /collections/travelers/documents/search?q=johnson
  &query_by=_id,first_name,last_name,email_address,employee_id
  &filter_by=company_group_id:=803f6614-a459-3c92-9375-fffeeeaa
  &highlight_fields=none
  &per_page=250
  &max_candidates=4
  &use_cache=false
  &search_cutoff_ms=1000

And the response, where

search_cutoff

was not applied and the search time is too high to be acceptable in our use case:

{
    "facet_counts":[],
    "found":803,
    "hits":[
    ...
    ],
    "out_of":3685109,
    "page":1,
    "request_params":{
      "collection_name":"travelers",
      "first_q":"johnson",
      "per_page":250,
      "q":"johnson"
    },
    "search_cutoff":false,
    "search_time_ms":4063
  }

I know that cutoff is not guaranteed, but I've tried different parameters, different values, and cutoff was not applied. That particular CompanyGroupId is one of the bigger ones, with around 100K travelers. Typesense version is

0.25.2

. Is there something I'm not considering? Any help is appreciated! Thanks!