Getting your cloud usage data
How to export your cloud usage data from AWS, GCP or Azure.
Having your cloud usage data exported to storage you own is a prerequisite to enriching it with Greenpixie. That enrichment can happen two ways. You can use Greenpixie's managed cloud usage report enrichment service, or the Greenpixie API, where you enrich your own usage data in your own environment. Either way, you first need the export.
This guide covers how to create that export on each of the big three platforms, Amazon Web Services, Google Cloud, and Microsoft Azure. Pick the guide for your platform, or work through more than one if your estate spans several.
What you are creating
On each platform you set up a recurring, provider-native export of billing and usage data, delivered to a bucket or container you own. The export contains service and usage detail, quantities, timestamps, regions, and account or resource IDs.
The data stays in your own environment. What you do with it next is your choice, covered in Next steps.
If you plan to use Greenpixie's cloud usage report enrichment service
The export mechanics are the same whichever route you take, but Greenpixie's managed cloud usage report enrichment service has a few preferences worth setting at export time:
- Format: Parquet is preferred, with zstd compression ideal. CSV with gzip is also accepted.
- File size: aim for no single file larger than 500 MB.
- Partitioning: each platform guide includes the required directory structure for delivery to Greenpixie.
None of this is required if you plan to enrich your own data through the API instead.
Considerations for partners
If you resell or manage cloud usage on behalf of multiple customers, and you will be using Greenpixie's cloud usage report enrichment service, add a further partition for a unique customer identifier, such as org_id, above the platform level. This keeps each customer's data cleanly separable through enrichment, whichever cloud providers they use.
<bucket_name>/<input|output>/<org_id>/<platform>/<account_id>/<timestamp>/{partition files}The <timestamp> directory identifies a single upload, in ISO 8601 basic format (UTC): YYYYMMDDTHHMMSSZ, for example 20260722T120000Z.
Keep the identifier consistent across exports and platforms so the same customer reconciles the same way everywhere. Confirm the exact layout with your Greenpixie contact.
Standard customer enrichments run daily between 01:00 and 04:00 UTC, processing the month-to-date. Partner enrichments instead run on demand: you signal that an upload is complete, and Greenpixie processes it from there. See Partner trigger, and Connect your data for cloud usage report enrichment.
Choose your platform
Get your AWS usage data →
Create a CUR 2.0 export from the management (payer) account to an Amazon S3 bucket.
Get your GCP usage data →
Enable the detailed billing export to BigQuery, then deliver it daily to a Cloud Storage bucket.
Get your Azure usage data →
Create a Cost Management (FOCUS) export at billing-account scope to a Blob Storage container.
Next steps
Once your export is running, you can enrich it with Greenpixie in two ways.