Connect a Postman API key to check your work
You do each step in Postman, then LiftOff uses your key to look at your workspace and confirm it's done. Connect once and it works for every module. Your key is saved in this browser and is only used to check your work in Postman.
Before anything can be queried, the data has to live somewhere Postman can reach. This part builds a dedicated workspace and your first dataset from two very different files, a CSV and a JSON export.
Create Your Datasets Workspace
Everything in this module, the dataset, the collection and the mock, lives in one workspace so the pieces stay together and easy to find.
-
Open the Postman desktop app. If you don't have it yet, download Postman and sign in. Datasets need a Solo, Team or Enterprise plan.
-
Click Workspaces in the top navigation → Create.
-
In Workspace name, enter [Your name] - Datasets (for example Alex - Datasets).
text- Datasets -
For Who can access, choose Only you and invited people.
-
Leave the template on Blank workspace, then click Create Workspace.
Your new workspace opens automatically with an empty sidebar.
Troubleshooting: If validation fails, check that the name ends with - Datasets (with the hyphen) and that you created the workspace yourself rather than joining someone else's.
Create Your First Dataset From a CSV
A dataset is a named container of data sources (where the data comes from) and views (SQL queries that decide what you see). You'll start with one source: the account list.
-
Save this as
accounts.csvon your computer. The header row names become your column names:accounts.csv5 rows · 4 columns Local fileaccountId company plan region 1 Northstar Logistics enterprise EMEA 2 Bluepeak Energy business AMER 3 Cobalt Health enterprise APAC 4 Lumen Retail starter AMER 5 Vertex Labs business EMEA -
In your Datasets workspace, click + in the left sidebar and select Dataset.
-
Name the dataset:
textbilling-dataAdd a short description if you like.
-
Choose a source type: Local File keeps the file on your machine, while Postman Cloud Store uploads it to Postman. Pick either, then choose
accounts.csv. You can also drag and drop it into the workbench. -
Preview the data and click Save.
Postman reads CSV, JSON and spreadsheet files (.xlsx, .xls and .ods), and there is no file size limit like the one data files have. Open the dataset and look at the views: Postman created one automatically that selects everything from your source.
Expected outcome: A billing-data dataset with one data source and one auto-generated view that returns all five accounts.
Add the Invoices as a Second Source
Real data rarely lives in one place. A dataset can combine many sources, even of different types, and later you'll join them with one query.
-
Save this as
invoices.json. It's a JSON array with one object per invoice:invoices.json7 rows · 5 columns Local fileinvoiceId accountId product amount status INV-1001 1 Platform Annual 12000 paid INV-1002 2 API Gateway 4800 paid INV-1003 1 Premium Support 2500 refunded INV-1004 3 Platform Annual 12000 paid INV-1005 4 Starter Plan 600 void INV-1006 3 Analytics Add-on 3000 refunded INV-1007 5 API Gateway 4800 paid -
If
billing-datais still open in the active tab from Step 2, stay there. Otherwise open it from the sidebar. Then click Add source. Select the Local File source type again (or Postman Cloud Store, to match what you chose in Step 2), chooseinvoices.json, preview the data and click Save. -
Each source now has its own auto-generated view, such as
source_accounts_default_view, listed under Datasets in the left sidebar. Postman names every sourcesource_plus the file name, so you now havesource_accountsandsource_invoices. These names are the tables you query with SQL in the next part.
Files aren't the only option. Here is everything a dataset can read from:
| Source | Where | Plan |
|---|---|---|
| CSV, JSON, spreadsheet | Local file or Postman Cloud Store | Solo, Team, Enterprise |
| MySQL, Postgres, SQL Server | Live database (optionally via an SSH tunnel) | Team, Enterprise |
| Custom JDBC driver | Oracle, Snowflake, BigQuery, Redshift, DuckDB and more | Enterprise |
Database credentials can be stored as vault secrets, with Link to vault secret on the connection fields, so they never sit in plain text.
Expected outcome: Your dataset has two sources and two auto-generated views.