Service Level calc from /api/v2/analytics/queues/realtime not matching dashboard

I’m trying to calculate Service Level % for our support queue using the real-time analytics API. The dashboard shows 92% SLA, but my script keeps returning 85%. I’m using the GET /api/v2/analytics/queues/realtime endpoint with the groupBy parameter set to queue.

Here is the relevant part of the JSON response I’m parsing:

{
 "groups": [
 {
 "key": {
 "id": "abc-123-def"
 },
 "values": {
 "queue": {
 "id": "abc-123-def",
 "name": "Support"
 }
 },
 "metrics": {
 "offeredCount": 100,
 "answeredCount": 90,
 "answerTime": {
 "mean": 12.5,
 "percentile90": 18.0
 },
 "waitTime": {
 "mean": 8.2,
 "percentile90": 15.0
 }
 }
 }
 ]
}

I assumed SLA is just answeredCount / offeredCount. That gives me 90%. But the dashboard says 92%. The docs mention serviceLevelPercentile but I don’t see that field in the real-time metrics object. Am I missing a specific metric? Or is the dashboard calculating it differently based on the 20-second threshold? I’ve tried using percentile90 from waitTime but that doesn’t seem right either. Just need to get the numbers to match for my adherence report. I’m in US/Central so maybe there’s a timezone offset issue with the interval data? No, this is real-time. Any code examples for the correct formula would be great. I don’t want to guess.

you’re mixing up the endpoint? that analytics path is for historical buckets, not real-time state. try /api/v2/interactions/queues for live metrics.