So we’ve been looking at the auto-summarization feature in Architect - specifically, trying to get the conversation summary data exposed for reporting. The idea is to avoid having to build a custom summarization pipeline, which, honestly, is a pain. We’re on Genesys Cloud, using Architect flows with the “Summarize Conversation” action and the default settings. Flows are published, and agents are using it, but the summary field isn’t populating in the conversation details view.
It’s not the end of the world, but now we’re wondering if there’s a way to pull these summaries via the analytics API. I’ve been banging my head against the /api/v2/analytics/conversations/details endpoint, trying to figure out if a new field was added for this, but nothing. The documentation is predictably unhelpful. We’re trying to avoid polling the entire transcript - pagination is already a nightmare, and that’s just adding latency.
I built a quick Python script using httpx to just dump all the available fields for a few conversations, and it’s all the standard stuff - wrapup codes, agent ID, queue time - nothing even close to a summary. Here’s the basic query:
import httpx
gc_url = "https://api.genesyscloud.com/v2/analytics/conversations/details"
access_token = "YOUR_TOKEN" # obviously replace this
headers = {
"Authorization": f"Bearer {access_token}",
"Content-Type": "application/json"
}
params = {
"queryType": "SIMPLE",
"timeUnit": "MINUTE",
"startDate": "2024-10-26T00:00:00.000Z",
"endDate": "2024-10-26T23:59:59.999Z",
"state": "ENDED",
"maxResults": 5
}
response = httpx.get(gc_url, headers=headers, params=params)
if response.status_code == 200:
data = response.json()
for conversation in data["results"]:
print(conversation.keys())
else:
print(f"Error: {response.status_code} - {response.text}")
We’re hoping to ship this reporting dashboard by Friday, and if we have to roll our own summarization, it’s going to be a long night. Is the API just lagging, or is there another way to access this data?