Getting started
Quickstart
This page takes you from an empty account to a base your team can work in. Budget five minutes.
1. Create a base
From the dashboard, New base. A base is a workspace: it holds several tables, its own views, automations and agents. Underneath, it is a dedicated PostgreSQL schema.
Three ways to start:
- Empty table — define the fields yourself.
- Import — CSV, Excel, or paste straight from a spreadsheet. Field types are inferred from the data, and you can correct them before confirming.
- Plain-language description — describe what you track (“a sales pipeline with contacts, companies and deals”) and the assistant proposes a schema. Nothing is created until you accept the proposal.
2. Get your field types right
The field type is what separates a database from a spreadsheet. Spend a minute on it — formulas, filters and automations all depend on it.
| What you need | Field type |
|---|---|
| One value from a fixed set | Single select |
| A link to another table | Linked record |
| A value pulled from the linked table | Lookup |
| An aggregate over linked records | Rollup |
| A per-row calculation | Formula |
All 28 types are available on every tier, self-hosting included.
3. Open a second view
A view is a reading of the table, not a copy. Switching views duplicates nothing.
On a table with a status field, add a Kanban view grouped by that status. On a table with dates, add a Calendar or Gantt view. Filters, sorts and groups belong to the view: a colleague’s “My overdue items” view does not change yours.
4. Invite your team
Share, top right. Permissions are set at base level (owner, editor, commenter, viewer) and can be narrowed per table and per field.
Editing is collaborative in real time: two people can work in the same table at once, with cursors and changes visible immediately.
Next
- Tables and fields — the field types in detail.
- Automations — trigger actions on your data.
- AI agents — delegate recurring work.