Pipelines
A pipeline is an ordered list of queries with its own cron (and its own timezone, like schedules). Each step says what to do on failure: stop the rest or carry on. History keeps every run with the steps as they were at the time, not as they are today. Available on Team and Business.
- Notify on failure (default), on success or always — to fixed addresses or to a broadcast list.
- A run in progress blocks the next one: `RunningSince` is what keeps two firings of the same pipeline from overlapping.
Syncs
A sync copies the result of a query to an external destination on the schedule you set. Three destinations today: webhook (POST with the rows as JSON), Google Sheets (one tab, replaced on every run) and Slack (a message with the table).
A sync passes through column masking before writing. There is no external destination that receives the raw value — the same reasoning as scheduled deliveries.
Notebooks
SQL cells executed in order, each with its result kept alongside. They are for investigating — what you want to keep becomes a saved query. Execution is asynchronous and each notebook's run history lives on its own page.