Setting up BigQuery for Warehouse Sync

Create a BigQuery dataset in your own Google Cloud project, let Retrixus load it, and connect it to Claude. About twenty minutes.

With Warehouse Sync, Retrixus loads your Recurly data into your own BigQuery every hour. The data stays in your Google Cloud project, under your billing and your access controls.

Warehouse Sync is an add-on to the Business plan, $49 a month or $468 a year. An owner adds it under Admin, Billing. Until it's added, Retrixus won't save a Warehouse Sync setting.

You need someone who can create a Google Cloud project, or use one you already have.

1. A Google Cloud project with billing

1. Go to the Google Cloud console and create a project (or pick an existing one).
2. Make sure it has a billing account. Storage for a few years of a busy Recurly site is usually a few dollars a month. Retrixus's loading costs you nothing: it runs in Retrixus's project.
3. Note the project ID (for example my-company-data).

2. A dataset for Recurly

1. Open BigQuery in that project.
2. Create a dataset, for example recurly, in the location you prefer (US is fine).

3. Let Retrixus load it

1. Open the dataset, then Sharing, Permissions, Add principal.
2. Principal: retrixus-api-prod@retrixus-mcp.iam.gserviceaccount.com
3. Role: BigQuery Data Editor.

That lets Retrixus write tables in this one dataset, and nothing else in your project.

4. Turn it on

In the Retrixus portal, under Admin, History: choose BigQuery, refreshed hourly, tick Retrixus loads it, and enter the project ID and dataset.

Within the hour, tables appear: recurly_accounts, recurly_subscriptions, recurly_transactions and recurly_invoices. Each shows the latest version of every record, with the full Recurly record in its record column. The first load of a large site's history can take a few hours, spread across hourly runs.

5. Connect it to Claude

In Claude, add the Google Cloud BigQuery connector under Settings, Connectors, and sign in with a Google account that can read the dataset. Then ask Claude about history: it uses BigQuery for that, and Retrixus for the latest hour.