Explore and store data

Flow gives you two places for data: Data Explorer, a SQL client for your database connections, and the Data Store step, which saves and finds records from inside an automation without any SQL.

Data Explorer

Data Explorer queries and manages your databases. In Flow, open Setup › Explorer. In the full Designer view it is Workspace › Operate › Data Explorer.

It works with database connections: PostgreSQL, MySQL, MongoDB and Redshift. To add one, see Connect a database.

Run a query

1. Open Setup › Explorer.

2. In Select a database connection…, pick a connection. Managed databases are listed under Platform databases, your own under Your connections.

3. Browse the tables in the schema tree on the left. Hover a table and select Preview data to see its first 100 rows.

4. Write SQL in the SQL Query editor and select Run Query (Ctrl+Enter or Cmd+Enter).

5. To download the results, select Export CSV.

A query returns at most 1,000 rows and stops after 60 seconds. Longer results are marked “(truncated)”.

History at the bottom lists your recent queries. They are kept in your browser, up to 50 per organization.

Who can read rows in Platform databases

Your environment’s built-in data store is one of the Platform databases. It holds your Apps’ people’s records as well as what your automations save. So rows in Platform databases are for your organization’s Owner and Admin only:

• An Owner or Admin of the whole organization can run SQL and Preview data. Each query is recorded in the audit log: who ran it, which tables, how many rows came back, and which Apps’ people were in the result. The rows themselves and the values in the query aren’t recorded.
• Anyone else, including an admin of one environment only and every API key, still sees the tables and their columns. A query or preview is refused: “This is your environment’s built-in data store, which holds your Apps’ people’s records. Only an organization owner or admin can query its rows; every query is audited.”

Your own database connections, under Your connections, aren’t affected.

Create a table

Select New Table above the schema tree. Enter the Schema, Table Name and Columns (Add Column for each), and turn on Show SQL Preview to see the statement. Select Generate & Run.

Danger

Queries run with the connection’s own database permissions, so DELETE, TRUNCATE and DROP work if that account allows them. The table menu’s Truncate Table and Drop Table permanently delete data. Before you confirm, make sure the right connection and table are selected, or write the statement yourself in the editor and read it before you run it.

Turn on a managed database

If you don’t have a database of your own, work.studio can run one for you. Select New Database next to the connection picker. In the Platform Databases window, choose an engine (PostgreSQL, MySQL, MongoDB or Amazon Redshift) and select Activate.

The database then appears under Platform databases, ready to use from Data Explorer and from automations.

Store records from an automation

The Data Store step saves and finds structured records with no SQL. It is in the step palette’s Logic category, and works the same in both views.

Save each new lead to a collection, then use the stored records later.

Records live in collections, such as leads. Pick what the step does under Operation:

Operation	What it does	Fields
Define collection (schema)	Creates a collection and its fields.	Collection, Fields (a JSON array of { name, type, required, description }), Store connection (optional)
Save record (insert / upsert)	Adds a record, or updates it if the Upsert key (optional) matches.	Collection, Record (JSON)
Find records (query)	Returns records that match.	Collection, Filters (JSON, equality on declared fields), Limit (optional)
Get record by id	Returns one record.	Collection, Record id
Update record by id	Changes one record.	Collection, Record id, Record
Delete record by id	Removes one record.	Collection, Record id
List collections	Lists the collections.	—
Describe collection	Returns a collection’s schema.	Collection

The Record JSON can use expressions, for example { "email": "{{input.email}}" }. Result variable (optional) stores the step’s result in an automation variable.

By default, records are kept in your environment’s built-in store (“Environment default (auto-provisioned store)”). To keep a collection in your own database, pick a database connection under Store connection (optional) when you define it.

Note

There is no screen for browsing data store collections. To look at the records, add a Find records (query) step and read its output in the test panel.

ConnectionsAdd a database connection.

DashboardsChart a SQL query on a dashboard.

ExpressionsUse values from the trigger and earlier steps.