List of conversation IDs per queue for inactive calls

[eastonce]

Hello, how can I get a list of conversation Ids based on past time/date interval, by queue and media type? I do not want an aggregate I already know how to get that. But instead, I need a list of conversation Ids. In other words, I have the target date interval, queue ID, and media type and I want to output a list of conversation Ids based on those filters.

Thanks.

[Eos_Rios]

An aggregate query\job would still be the only way to achieve that.

[eastonce]

Hi thanks for the reply. Can you suggest which one?

[Eos_Rios]

POST /api/v2/analytics/conversations/aggregates/query

Rough example;

{
  "interval": "2023-09-13T00:00:00/2023-09-14T00:30:00",
  "granularity": "P1D",
  "timeZone": "America/Chicago",
  "groupBy": [
    "conversationId"
  ],
  "metrics": [
    "nConnected"
  ],
  "flattenMultivaluedDimensions": true,
  "filter": {
    "predicates": [
      {
        "dimension": "queueId",
        "value": "your-queue-id",
        "type": "dimension"
      },
      {
        "type": "dimension",
        "dimension": "mediaType",
        "operator": "matches",
        "value": "voice"
      }
    ],
    "type": "and"
  }
}

[eastonce]

thank you. that works. however, i should have been more specific. i am trying to also filter for conversations where channel = external and for a specific topic id from speech and text analytics. i have been using POST
/api/v2/analytics/transcripts/aggregates/query to generate a count using the filtering criteria i mentioned above. however, i want to confirm that i am only querying for topics from speech and text being captured from the external channel (customer). Unfortunately, the api you suggested above does not allow me to filter by topic id or channel. Would any of the transcripts API allow me to do this?

[Eos_Rios]

Wouldn't you just add;

,
      {
        "type": "dimension",
        "dimension": "purpose",
        "operator": "matches",
        "value": "customer"
      }

to the example I gave to achieve that?

or

value": "external"

or you could do an OR around both.

[eastonce]

that returns empty when i add that as an additional predicate. but even if it wasnt empty, how can i filter for a specific topic within that interaction's speech? here's what i have so far that returns blank:

{
"interval": "2023-09-25T21:31:16.150Z/2023-09-26T22:06:21.193Z",
"granularity": "P1D",
"timeZone": "America/Chicago",
"groupBy": [
"conversationId"
],
"metrics": [
"nConnected"
],
"flattenMultivaluedDimensions": true,
"filter": {
"type": "and",
"predicates": [
{
"dimension": "queueId",
"value": "1f1fed5a-68ef-4a32-9a15-84e3646440c5",
"type": "dimension"
},
{
"type": "dimension",
"dimension": "mediaType",
"operator": "matches",
"value": "voice"
},
{
"type": "dimension",
"dimension": "purpose",
"operator": "matches",
"value": "external"
}
]
}
}