We are attempting to construct an Analytics API aggregation query that groups interaction data by both queue and media type. The goal is to retrieve a summary of handle times and wait times for specific queues across voice and webchat channels. The request is made to POST /api/v2/analytics/conversations/details/query.
The query body includes the necessary filters for the date range and the specific queue IDs. The grouping structure is defined as follows:
{
"groupBy": [
{
"property": "queue.id"
},
{
"property": "mediaType"
}
],
"metrics": [
{
"name": "handleTime",
"type": "sum"
},
{
"name": "waitTime",
"type": "sum"
}
]
}
The API returns a 200 OK status, which indicates the request syntax is valid. However, the data array in the response payload is consistently empty. We have verified that interactions exist for the specified queues and date range by running a standard details query without the grouping parameters. That query returns the expected list of interactions.
The issue appears to be isolated to the aggregation logic. When we remove the groupBy section and only request the metrics, the response contains a single object with the total sums for all interactions. This confirms the metrics themselves are accessible. The problem arises only when attempting to group the data.
We have tried using queue.name instead of queue.id, but the result remains the same. The documentation suggests that grouping by mediaType is supported, but combining it with queue.id might have specific requirements or limitations that are not clearly documented. We are using the Genesys Cloud Python SDK version 3.1.0 to make these requests. The code snippet for the SDK call is straightforward:
response = api_instance.post_analytics_conversations_details_query(body=query_body)
The environment is a Genesys Cloud instance in the Europe region. The timezone for the queries is set to UTC. We need to understand if there is a specific format or additional filter required to successfully group by these two properties. Any insights into the correct structure for this type of aggregation query would be helpful.