Skip to main content
This example moves public.orders into analytics.orders_live.

​
1. Prepare the destination

Create the destination table before configuring the pipeline: 
CREATE SCHEMA IF NOT EXISTS analytics;

CREATE TABLE IF NOT EXISTS analytics.orders_live (
  id bigint PRIMARY KEY,
  customer_id bigint,
  order_status text,
  total_amount numeric,
  created_at timestamptz,
  updated_at timestamptz
);

​
2. Create connections

Create and test:
  • PostgreSQL source connection with read access to public.orders
  • PostgreSQL destination connection with write access to analytics.orders_live

​
3. Create the pipeline

  1. Open Data Pipelines.
  2. Create a new pipeline named Orders to Analytics.
  3. Select the source connection.
  4. Open the builder.

​
4. Configure Source

  1. Open the Source node.
  2. Discover schema or refresh tables.
  3. Include public.orders.
  4. Preview source rows.

​
5. Configure Destination

  1. Open the Destination node.
  2. Select the destination connection.
  3. Keep sync mode as FULL_TABLE for the first run.
  4. Keep write mode as Upsert.
Optional Normalisation rules:
RuleTableColumnTarget
Renamepublic.ordersstatusorder_status
Castpublic.orderstotal_amountnumeric

​
6. Write the SQL model

In the SQL Layer tab, set final delivery table to analytics.orders_live and use: 
SELECT
  id,
  customer_id,
  order_status,
  total_amount,
  created_at,
  updated_at
FROM {{ source('raw', 'public__orders') }}
WHERE order_status IS NOT NULL
Validate SQL, preview output, and save the destination.

​
7. Run and verify

Run manually from the Destination node. Then check run history for:
  • success status
  • rows written
  • delivery target analytics.orders_live
  • warnings
Query the destination table directly if you want an external verification.

​
8. Move to Incremental

After the manual Full Table run succeeds:
  1. Switch sync mode to INCREMENTAL.
  2. Enter replication key updated_at.
  3. Save.
  4. Run manually again.
  5. Add a schedule.