Quickstart

Quickstart

1. Sign in

Open console.thunderduck.ai (opens in a new tab) and sign in with email or Google. Your company is created on first sign-in; colleagues you invite join the same company and see the same catalogs.

2. Connect a catalog

Catalogs → New catalog. Pick a type, give it a lowercase name (letters, digits and underscores), and fill in the connection details.

The name matters: it becomes the first part of every table reference, so sales_pg means you write sales_pg.public.orders.

Credentials are verified before anything is saved — a wrong password fails at this step rather than surfacing later inside a query. They then go straight to a secrets manager and are never stored in our database or returned by any API.

3. Run a query

Studio, then:

SELECT * FROM sales_pg.public.orders LIMIT 100;

Studio shows the plan, the estimated cost, progress while it runs, and the rows when it finishes. Anything above roughly ten seconds of predicted runtime is handed to the background and polled, so a long query does not block the page.

4. Join across sources

The point of a catalog is that everything is queryable together. A Postgres table joined to an Iceberg table joined to a spreadsheet in object storage:

SELECT o.id, o.total, t.region, x.target
FROM   sales_pg.public.orders            o
JOIN   lake.default.territories          t ON t.code = o.territory
JOIN   read_xlsx('files/targets.xlsx')   x ON x.region = t.region
WHERE  o.created_at >= current_date - 30;

5. Keep a result

Results are scratch by default and expire after a day. Save keeps one for a year, and Download gives you a time-limited link to a Parquet file.

Next

IntroductionMigrating from BigQuery