Backend functions explained
What a backend function is in plain terms, why some features need one, and what you do and do not have to do.
Last updated September 17, 2026
On this page
Most of your app runs in the visitor's browser. Some work must not: using a key you hold to talk to another service, a job that runs at night, a webhook from a payment provider. That work runs as a backend function, on the server, where visitors cannot see or change it.
What you have to do
Usually nothing. You describe the feature and the builder decides it needs a function, writes it, tests it and switches it on. If the function needs a key, the builder asks for it with a card.
Where you see them
- Automations: pick which function an automation runs.
- API: each function can be called from your own server.
- Logs: what each run did.
Running as the app
A scheduled job has no signed-in user. To read or write rows no visitor could, a function runs as the app itself. The builder grants that to a function it wrote and tested and tells you so.
Developers can read the full function reference.