Cloud cost management
This section is about FinOps - what you spend with your cloud providers (AWS, Hetzner, Azure, GCP, and the rest) - not about your bill from Sencai. Those are entirely different things: your provider bill is money that moves between you and, say, Hetzner; your Sencai bill is covered in Billing & plans. Sencai observes and analyzes the first one; it doesn’t touch the second on this screen.
Where the numbers come from
Section titled “Where the numbers come from”Cost data in Sencai comes from two places:
- Your connected cloud providers’ own billing data. Once you’ve connected a cloud account, Sencai periodically pulls cost and usage data directly from that provider’s billing API and stores it as cost records.
- Sencai’s own resource records - the instances, networks, and other managed resources it already tracks for you, used to attribute spend to a specific resource, service, or region.
Not every connected provider necessarily has cost ingestion built for it yet. If a provider you use doesn’t show up in the cost breakdown at all, treat that as “not yet supported” rather than “nothing to report” - check with your account team, and see Connect a cloud provider for the current provider list.
The Cost Records screen
Section titled “The Cost Records screen”Infrastructure → Cost Records (/gravity/costs) is the home page for cost data. It has two
views:
- FinOps overview - total spend for the organization, a breakdown by provider, a spend trend chart over a 30/60/90-day window you can toggle, and a monthly budget indicator if your account team has configured one for your organization. A simple statistical check flags a day whose spend is unusually far outside your recent average as a possible anomaly.
- Cost records (shown as “COGS (FOCUS v1.3)”) - the underlying line-item data, in the format of FOCUS, an open specification for standardizing cloud cost and usage data across providers. Pick a period (this month, last month, or the last 90 days) and you get four summary figures (total spend, top provider, top category, top service), a per-provider breakdown, a per-category breakdown, and an expandable table of the individual records behind them - resource ID, service, region, cost, and billing period. You can export the current period’s records as CSV.
Both views are scoped to the organization you currently have selected - switch organizations from the picker at the top of the screen the same way you would anywhere else in the app.
Granularity and refresh cadence
Section titled “Granularity and refresh cadence”Cost records are ingested at whatever granularity your provider’s own billing API reports - in practice, per resource per day. Ingestion runs on a recurring schedule that your account team controls; it isn’t real-time; a change to your infrastructure today won’t necessarily be reflected in cost data until the next ingestion cycle. If you need to confirm the freshness of a specific number for an audit or reconciliation, ask your account team what your organization’s ingestion cadence is.
Currency
Section titled “Currency”Each cost record carries the currency your provider reported it in. The FinOps overview’s totals use the currency present on your organization’s records - most commonly EUR - and don’t automatically convert between currencies if your account has records in more than one. If you run infrastructure across providers billing in different currencies, treat totals as a same-currency roll-up rather than a normalized figure.
What’s included by plan
Section titled “What’s included by plan”The depth of FinOps tooling - anomaly detection, forecasting, multi-cloud comparison, and custom reporting - differs by organization tier. See Billing & plans for what’s included at each level; this documentation describes how each feature behaves when it’s available to you, not which plan unlocks it.
What’s next
Section titled “What’s next”- Cost allocation - attribute spend to teams, projects, or environments
- Right-sizing - recommendations for over-provisioned instances
- Forecasting & anomalies - projected spend and unusual-spend detection
- Spend caps - configure soft and hard organization spending limits
- Connect a cloud provider - required before any cost data appears