Uniphore Customer Portal

A filter defines what type of data you want to filter by (fieldName), what condition the filter should regard as a match (operator), and what data value you want to filter for (stringValue), for filters using a fieldName that requires a date and time to be specified there's also timestampValue - using these parameters you can apply your request to or get conversations with specific metadata values, for example you could find conversations where the agent's extension contains "384":

"groupFilter": {
    "leftFilter": {
        "fieldName": "agent.extension",
        "operator": "CONVERSATION_FILTER_OPERATOR_CONTAINS",
        "stringValue": "384"
    }
},

Group filters use the logic above but can combine multiple filters to match conversations with multiple conditions, for example where more than one data type matches specific conditions, or the same data type matches two specific conditions. Group filters combine filters using AND/OR operators, and use leftFilter and rightFilter array parameters to differentiate them.

The simple graphic below illustrates how each of them work. In this example you have three searches – A, B, and C.

For example, you could apply the request to conversations to/from the Sales Team AND Pre-Sales Team:

"groupFilter": {
     "leftFilter": {
        "fieldName": "agent.department",
        "operator": "CONVERSATION_FILTER_OPERATOR_EQUAL",
        "stringValue": "Sales"}
    },
    "operator": "CONVERSATION_LOGICAL_OPERATOR_AND"
    "rightFilter": {
        "fieldName": "agent.department"
        "operator": "CONVERSATION_FILTER_OPERATOR_EQUAL"
        "stringValue": "Pre-Sales"
    }
},

Group filters can also contain other group filters for even more precision and flexibility, this works by nesting group filters. Whenever you group a set of filters together using AND/OR, you’re basically saying search for all these things first and provide the result, then combine the result with any other filters that you have defined further down in the groupFilter.

Nested group filters use left and right parameters to differentiate them, similar to group filters, but use leftGroupFilter and rightGroupFilter to indicate that they are nested. In the example below you can see that groupFilter contains a leftGroupFilter, a rightGroupFilter and an operator - as these are nested you can see that leftGroupFilter and rightGroupFilter each contain an operator and another leftGroupFilter / rightGroupFilter pair each with their own leftFilter.

Following this nested logic you can create a very precise filter - in the example below, I filtered for all calls from two agents using OR, then grouped a department filter (Sales) and a Call Direction filter (CONVERSATION_DIRECTION_OUTGOING) with AND to give me all Outgoing calls from the Sales department, using OR at the top means my results would provide all calls by either agent (Agent 384713 OR Agent 606939) as well as any outgoing sales calls.

Logical expression: (agent.id EQUAL 384713 OR agent.id EQUAL 606939) OR (callProperties.department EQUAL sales AND callProperties.callDirection EQUAL CONVERSATION_DIRECTION_OUTGOING)

"groupFilter": {
    "leftGroupFilter": {
        "leftGroupFilter": {
            "leftFilter": {
                "fieldName": "agent.id",
                "operator": "CONVERSATION_FILTER_OPERATOR_EQUAL",
                "stringValue": "384713"
            }
        },
        "operator": "CONVERSATION_LOGICAL_OPERATOR_OR",
        "rightGroupFilter": {
            "leftFilter": {
                "fieldName": "agent.id",
                "operator": "CONVERSATION_FILTER_OPERATOR_EQUAL",
                "stringValue": "3606939"
            }
        }
    },
    "operator": "CONVERSATION_LOGICAL_OPERATOR_OR",
    "rightGroupFilter": {
        "leftGroupFilter": {
            "leftFilter": {
                "fieldName": "callProperties.department",
                "operator": "CONVERSATION_FILTER_OPERATOR_EQUAL",
                "stringValue": "Sales"
            }
        },
        "operator": "CONVERSATION_LOGICAL_OPERATOR_AND",
        "rightGroupFilter": {
            "leftFilter": {
                "fieldName": "callProperties.callDirection",
                "operator": "CONVERSATION_FILTER_OPERATOR_EQUAL",
                "stringValue": "CONVERSATION_DIRECTION_OUTGOING"
            }
        }
    }
},

fieldName Values

As mentioned above, fieldName is the type of metadata you want to filter by - valid options, what they mean, and any specific conditions are detailed in the table below.