Data Tables

Data Tables give you first-class relational storage inside ByteChef. Create a table with typed columns, manage rows in a spreadsheet-like grid (or by importing a CSV), and read or write that data from your workflows — without standing up an external database.

Use them for lookup tables, configuration, queues, feature flags, deduplication keys, or any small structured dataset your automations need to share.

Data Tables are workspace-scoped and per-environment: each table belongs to a workspace, and the rows you see depend on the selected environment (Development, Staging, Production).

---

Create a table

1. In the Automation workspace, open Data Tables from the sidebar.
2. Click New Table.
3. Enter a name and an optional description.
4. Define at least one column — give it a name and a type (see below).
5. Click Create.

Every table automatically has an id column that uniquely identifies each row, so id is reserved and can't be used as a column name.

Column types

Values are coerced to the column's type, so a row written from a workflow or imported from CSV is stored in the correct shape.

---

Manage columns

Open a table and use its column menu to:

• Add a column — pick a name and type.
• Rename a column.
• Delete a column — removes that column's values from every row.

---

Manage rows

The table opens in a spreadsheet-like grid where you can:

• Add a row — fill in the new empty row at the bottom.
• Edit a cell — click it and type; changes save when you move off the cell. Each cell uses an editor matched to its column type (text, number, checkbox, date, or date-time picker).
• Delete rows — select one or more with the row checkboxes and delete them.
• Import a CSV — from the grid's ⋮ actions menu, bulk-load rows from a file whose header matches your column names.
• Export a CSV — from the same ⋮ menu, download the current rows.

Larger tables load more rows as you scroll. Use tags to organize tables once you have several.

On the Data Tables list page you can rename, duplicate, or delete a table from its row, and filter the list by tag from the left sidebar.

---

Use a table in workflows

Add the Data Table component to a workflow to read and write your tables as part of an automation. Select the target table, and the action's inputs adapt to that table's columns.

Actions

Triggers

A workflow can also start in response to a row change:

• Record Created — a new row was added
• Record Updated — a row was changed
• Record Deleted — a row was removed

This makes Data Tables a lightweight event source — for example, "when a row is added to the Customers table, send a welcome email."

---

Scope and access

• Workspace-scoped — a table is available to every project and workflow in its workspace.
• Per-environment — the rows are tied to the selected environment, so Development data stays separate from Production.
• Workspace membership — a table is reachable by the members of its workspace. There is no per-operation role split for data tables in the latest released version: a member who can open the workspace can read and write its tables. Finer-grained viewer/editor enforcement is on the upcoming release track.

---

Storage limits

The total size of all data tables in a tenant is capped — by default at 50 MB. As usage approaches the cap, a warning alert appears in the Data Tables UI; once the cap is reached, writes that would grow the data (inserting rows, importing CSV, updating rows) are rejected with a limit-exceeded error until space is freed.

Self-hosted administrators can change or remove the cap with the BYTECHEF_DATA_TABLE_MAX_SIZE_BYTES environment variable (0 = unlimited).

---

When to use a Data Table

A Data Table is a small relational store you get without provisioning anything. That makes it the right shape for some jobs and the wrong shape for others.

Patterns it fits

What a Data Table is not

• Not a transactional system of record. There is no cross-row transaction or multi-writer locking. If several systems write the same data concurrently, point the workflow at a real database through the matching component instead.
• Not a data warehouse. The bound to plan against is total stored size, not a row count — see Storage limits above. How many rows fit under a given size depends entirely on how wide they are, so size the data, not the row count.
• Not a search index. For matching by meaning rather than by column value, use the Knowledge Base.

Data Table or external database