https://api.bindbee.dev/api/hris/v1/. Twenty-one models, grouped by what they describe rather than by which endpoint returns them.
Every collection endpoint is paginated and takes the same filters for ids, remote_id and modified_after - see Pagination. Every request needs both an API key and a connector token - see Authentication.
The models
The split that catches people most often is Employee versus Employment. An employee is a person; an employment is one period of work. One person routinely has several - see Employee & org models.
What accepts writes
Four models. Everything else is read-only through the unified API.
Each has a Meta endpoint returning the request schema for that connector’s integration - build the body from it rather than hardcoding one, since required fields differ per integration. See Fetch a schema.
There is no update or delete on unified models. Where a correction is possible at all, it goes through Passthrough.
Where to start
Pagination
Every collection endpoint works this way.
Employee & org models
How the models relate, and where fields actually live.
Scope & permissions
Why a field you expected came back empty.