Get a single interaction (Conversation Analyzer API)

The GET /interactions/{guid} endpoint returns categorization and call time breakdown data for a single call recording. Optionally, it can also return the transcript of the recording.

Request parameters

Path parameters

guid

The unique identifier for the recording’s call.

Query string parameter

transcript

A flag indicating whether you want to receive the transcript information for the recording. Default is false.

Requests

Request the transcript, categorization and call time breakdown data for a single call

curl -L -X GET "https://***.api.newvoicemedia.com/conversation-analyzer/interactions/0164705c-c206-dce9-d61a-9899855326a0?transcript=true" 
-H "Accept: application/vnd.newvoicemedia.v3+json" 
-H "Authorization: <ACCESS_TOKEN>"

Responses

A response to a valid GET /interactions/{guid} request contains a single interaction item with information about the recording for a single call.

Successful responses

Successful requests return a 200 Success code.

{
    "categories": [
        {
            "category": "Politeness",
            "categories": [
                {
                    "name": "Greeting",
                    "matches": "1"
                },
                {
                    "name": "Closing",
                    "matches": "3"
                }
            ]
        },
        {
            "category": "Communication",
            "categories": [
                {
                    "name": "Responsive",
                    "matches": "1"
                }
            ]
        }
    ],
    "guid": "0164705c-c206-dce9-d61a-9899855326a0",
    "language": "en-us",
    "status": "processed",
    "confidence": "Low",
    "callTimeBreakDown": {
        "agent": 5,
        "customer": 1,
        "silence": 4,
        "crossTalk": 0
    },
    "transcript": "Case.\nHi, and yeah, just a call here. Thank you bye."
}

where an interaction item contains the following parameters:

• categories. Categories represent groups of defined phrases that an agent or customer uses during a call.
  • category. A single category represents a defined phrase type that an agent or customer uses during a call.
  • categories. (v2 and v3 only)
    • name.
    • matches.
  • subcategories. (v1 only)
  • name.
• guid. Unique identifier of the recording’s call. guid will match the unique identifier specified in the request.
• language. The ISO code for the language used when transcribing the call. (v3 only)
• status. The state of the end-to-end processing of the call recording. status is one of the following values:
  • processing. The recording is being processed.
  • processed. The recording has been processed.
  • failed. The processing of the recording failed.
• confidence. The level of confidence in the transcript. confidence is one of the following values:
  • high
  • medium
  • low
• callTimeBreakdown. The time, in seconds, that the recording included one of the following:
  • agent. Time when only the agent was talking.
  • customer. Time when only the customer was talking.
  • silence. Time when nobody was talking.
  • crossTalk. Time when the agent and customer were both talking.
• transcript. If the transcript request parameter was true, then transcript contains a transcript of the recording; otherwise, transcript is null.
• sentimentSummary. This is included in the response only when a given interaction has a sentiment summary available. (The sentiment feature must also be enabled when the interaction is being transcribed and categorized.)
• overallSentiment. Average sentiment score of the whole conversation.
• categories. Percentage of the call in each of five categories:
• veryNegative. 
• negative.
• neutral. 
• positive. 
• veryPositive. 
• distributionOverTime. An array of ten sections of of the interaction each representing 10% of interaction time and its average sentiment score.
• score. Average sentiment score in a given section of an interaction.
• span. Represents start and end percentage of a given section. start is inclusive, end is exclusive. For example, where start is 11 and end is 20, the range will include the sentiment score for between >=10% and <20%.
• start.
• end.