Datasets
Bring business data into OpsLyft so cloud cost can be joined to revenue, usage, teams, and customers.
A dataset is data that lives outside your cloud bill: revenue by customer, product usage events, org hierarchy, LLM token counts, vendor invoices. Once you map its account, team, and date columns, it joins to cloud cost, so you can build unit costs (cost per customer, per request, per token) and allocate spend.
When a dataset's first sync finishes, you can pick it in the widget builder and use it in Dimension Studio.
The datasets list
The header shows the total and how many need attention. Tabs split datasets into All, Yours, and Platform (maintained by OpsLyft).
| Column | Meaning |
|---|---|
| Dataset | Name and identifier, such as Revenue by customer, monthly · revenue_monthly |
| Source | Type and location, such as SQL database · postgres://db-prod.internal/finance |
| Status | Synced, Paused, Stale, and so on, with the time. See Dataset status. |
| Rows | Row count from the last sync |
| Owner | Who's responsible for it |
Datasets that need attention are listed first. Filter with the search box, Source, and Status, or select Export list.
Example datasets
| Dataset | Source | Used for |
|---|---|---|
| Revenue by customer, monthly | SQL database | Cost as a share of revenue, per customer |
| Customer usage events | Amazon S3 | Cost per active customer or per event |
| Org hierarchy | Google Sheets | Rolling teams up to departments for chargeback |
| Chargeback overrides | Google Sheets | Manual allocation exceptions |
| LLM token usage | Amazon S3 | Cost per token by team or feature |
| Region to entity map | File upload | Allocating cost to legal entities |
| Vendor invoices | File upload | Adding non-cloud vendors to the bill |
| Query volume | Prometheus | Cost per query |