Stage percentiles and ended reasons over the Calls of a window

GET /v1/analytics/latency

Stage percentiles, ended reasons and per-Agent timings over the caller Turns of the ended Calls of a window (ADR 0063). From 1.57.0 (workspace ADR 0048, observability-and-analytics.md section 5) it reads the latency rollups: window=custom with from and to reads a range, hourly rows up to 35 days and daily rows beyond, at most 760 days (range_too_long); the filters narrow it; group_by groups by the fixed dimensions of ANALYTICS_DIMENSIONS_BY_REPORT.latency, any other answering 422 validation_failed and the issue dimension_not_supported; and the answer adds every stage with p50, p90, p95 and p99 computed from bucket counts, its Turn count, whether it is significant and its adherence to the target, with the bucket counts under include_buckets=true. Verbose and test Calls are left out unless selected. A request of 1.56.0 answers as before; the route keeps calls:read.

Request

Set VOCAL_API to the base URL of your API and VOCAL_API_KEY to an API key with the calls:read scope (Authentication and keys). Replace each placeholder, such as agt_... or <name>, with your own value.

curl "$VOCAL_API/v1/analytics/latency" \
  -H "Authorization: Bearer $VOCAL_API_KEY"

const response = await fetch(`${process.env.VOCAL_API}/v1/analytics/latency`, {
  headers: {
    Authorization: `Bearer ${process.env.VOCAL_API_KEY}`,
  },
});
console.log(response.status, await response.text());

import os

import requests

response = requests.get(
    f"{os.environ['VOCAL_API']}/v1/analytics/latency",
    headers={
        "Authorization": f"Bearer {os.environ['VOCAL_API_KEY']}",
    },
)
print(response.status_code, response.text)

Parameters

Responses

Fields of the 200 answer

The shape is LatencyReport.