points.summary (version 1)
One summary per point over [from, to) (up to 380 days): count, min and max with their times, average, sum, and first and last reading with their times. Cheaper than points.timeseries when one number per point is enough, e.g. the warmest room last month.
bearerAuthorizationBearer <JWT>A signed-in Relynk user's access token: OAuth 2.0 authorization code with PKCE at https://auth.relynk.io/authorize and https://auth.relynk.io/oauth/token, with audience https://ai.relynk.io/builder. The caller sees exactly what that user may see.
tenantId?stringThe tenant to read. May be left out when the caller belongs to exactly one tenant; GET /me lists them.
uuidapplication/json- body
pointIds*array<>Point ids (from points.list), at most 500
1 <= items <= 500from*stringStart, inclusive: "YYYY-MM", "YYYY-MM-DD" (Oslo midnight) or ISO 8601 with offset, e.g. "2026-09-01T06:00:00+02:00"
to*stringEnd, exclusive: "YYYY-MM", "YYYY-MM-DD" (Oslo midnight) or ISO 8601 with offset, e.g. "2026-09-01T06:00:00+02:00"
The function output
application/json- response
timeZone*stringfrom*stringResolved start, Oslo time with offset
to*stringResolved end (exclusive), Oslo time with offset
points*array<>curl -X POST 'https://ai.relynk.io/builder/v1/points.summary/v1' \ -H 'Content-Type: application/json' \ -d '{ "pointIds": [ "077ab0e7-997d-450a-aa83-a0ee4c7f1644" ], "from": "string", "to": "string"}'{ "timeZone": "Europe/Oslo", "from": "string", "to": "string", "points": [ { "pointId": "string", "unitKey": "string", "count": 0, "min": 0, "minAt": "string", "max": 0, "maxAt": "string", "avg": 0, "sum": 0, "first": 0, "firstAt": "string", "last": 0, "lastAt": "string" } ]}points.raw (version 2) POST
Individual readings of up to 200 points over [from, to) (up to 31 days), ordered by point then time. Prefer points.timeseries or points.summary for analysis; use this to inspect exact readings. Pages: pass nextCursor back until it is null. Each page has one series per point that has readings in it, so merge pages by pointId.
points.timeseries (version 1) POST
Readings of up to 100 points aggregated into Oslo-time buckets over [from, to): one shared time axis t and one array per requested metric per point, null where a bucket has no readings. Days, weeks (Monday) and months start at Oslo midnight, so DST days have 23 or 25 hours. Limits: 15m buckets up to 62 days, others up to 400 days, and points × buckets up to 50 000; split longer ranges into several calls.