Getting started

Connect two nodes, and add what isn’t in your database

Connections are how a flow explains movement, not just membership. And because real systems don’t stop at the database, you can add the API, the nightly job, or the vendor feed that determines what the data actually means.

1. Hover a node and drag from its handle to another node to draw a connection.

   Direction is the convention that makes a flow readable at a glance: the left side is upstream (what feeds in), the right side is downstream (what it feeds).

2. An arrow always means data moving. So a scheduled job, an API, or a file drop belongs *in* the flow, where the data passes through it — not off to the side as a trigger. That a job runs a procedure is a fact for its description, not an arrow.

   If some arrows meant "data goes this way" and others meant "this triggers that," a reader would have to decode every one. Keeping a single meaning is what keeps the diagram readable.

3. To add context that isn’t in your schema, use Add Node and pick an Automation or External Systems type.

   The parts of a flow that aren’t in the schema often decide what it means — the job that populates a table, the API that consumes it.

4. Automation types: Cron / Job, Worker / Script, ETL Pipeline. External Systems types: API / Service, FTP, IO / File, Data Warehouse, BI / Dashboard, Queue / Event Bus.

5. Connect these to your objects the same way, then Save.