Connect to GCP for Recommendations
Here are the steps to manually set up your GCP Savings integration with Amberflo. You will configure a BigQuery export of data and grant Amberflo the necessary access. This setup builds directly on your GCP billing data integration that you did previously. You’ll reuse the same project, dataset, and IAM role you already created there.
Step 1. Reuse Your Project and Dataset
Use the same Project ID and Dataset ID that you created when setting up the FOCUS billing export. No need to create new ones.
Step 2. Configure Export in Recommendations Hub
- Go to the Recommendations Hub Export page.
- Choose BigQuery as the export destination.
- Select the same Project ID and Dataset ID you used for the FOCUS integration.
- Save the export.
Once enabled, GCP will start writing daily recommendations to your dataset. The first export usually appears within a couple of days.
Step 3. Verify IAM Role Permissions
Confirm that the custom IAM role you created for the FOCUS integration includes these permissions:
- bigquery.datasets.get
- bigquery.jobs.create
- bigquery.tables.export
- bigquery.tables.get
- bigquery.tables.getData
- bigquery.tables.list
This role should already be assigned to the Amberflo service account:
Step 4. Wait for Data to Arrive
After the first export runs (up to 2 days), you’ll see two tables appear in your dataset:
- recommendations_export
- insight_export
These are updated daily with fresh data.
Step 5. Share Setup Details with Amberflo
Send Amberflo the following:
- Project ID
- Dataset ID
Once you share this information, Amberflo can begin ingesting and analyzing your GCP savings recommendations.