Usage

Bring usage in

Open Usage → Import usage and choose a CSV file with these columns:

event_id,customer,meter,quantity,occurred_at
e-001,NW-1001,events,1000000,2026-07-02T10:00:00-07:00
e-002,NW-1001,events,2346778.5,2026-07-19T16:30:00-07:00

• event_id is your stable id for the event.
• customer is the Customer ID; meter is the meter's code.
• quantity is an exact decimal.
• occurred_at is ISO 8601 with an explicit offset.

Each file is one batch with its own report: Accepted, Duplicates and Rejected, row by row. An event id already recorded with the same values is a duplicate and changes nothing, so sending a file twice is safe. The same id with different values, an unknown customer or meter, or an unreadable time is rejected.

From a pipeline

A metering pipeline can post events directly with a usage credential (see People, roles and agents):

curl -X POST https://your-nectar.example/api/nectar/usage.submit \
  -H "Authorization: Bearer $NECTAR_USAGE_KEY" \
  -H "Content-Type: application/json" \
  -d '{"events":[{"event_id":"e-003","customer":"NW-1001","meter":"events","quantity":"250","occurred_at":"2026-07-20T09:00:00-07:00"}]}'

Up to 5,000 events go in one call, reported row by row the same way.

Check before rating

Before rating shows, for any month and each metered subscription:

• the period's boundaries in your billing timezone,
• each meter's event count, exact period total and what is already billed,
• the days without usage,
• events that arrived after the period ended,
• unbilled events from earlier periods (billed as late usage in the next period, at the price in effect when they occurred).

Correcting usage

Usage is append-only. To correct an event, an agent or a colleague proposes a usage correction that voids the wrong events and adds the right ones; Nectar checks the additions and computes the effect on the draft invoice before a person approves it (People, roles and agents).