Google Cloud

A claims-triage model on Vertex AI

Cloud cost, booked from the billing export through the initiative's budget, for September

$3,420 + $1,180 + $310

$4,910 booked to the claims initiative for September, from a $61,200 Google Cloud bill

AI Spend counts Vertex AI and the Gemini API as AI on their own, and the Cloud Run because the budget claims it. The rest of the bill is named as outside the AI figures, not added to them.

Illustrative figures. Roiva keeps every input with the entry it calculates, and none of it counts until it's approved.

Reads: Your budgets, and monthly cost by project and service from your Cloud Billing export in BigQuery. The first sync covers 12 months and later ones the last two. Labels only where a budget filters on them.

Keeps: Records of budgets, plus the totals worked out from them.

What Roiva keeps from each platform, field by field →

You connect Google Cloud from inside Roiva, so you need a Roiva workspace first. Sign in to connect it now. New to Roiva? Start free with your work email, take the free assessment and create your workspace at the end, or request an invite.

1. In Google Cloud: Open the Billing export page (console.cloud.google.com/billing/export), choose your billing account, and look at Standard usage cost on the BigQuery export tab. If it already names a project and dataset, the export is on and you can skip the next two steps
2. In Google Cloud: If it's off, enable it and pick the project and dataset to hold it. The project must be linked to this same billing account, so a project still in the free BigQuery Sandbox can't hold the export, and the dataset should have table expiration turned off. Enabling it needs the Billing Account Administrator or Billing Account Costs Manager role, plus BigQuery User on the project. A US or EU multi-region dataset also gets the previous month's costs
3. In Google Cloud: Give the export time to fill. Google creates the table shortly after the export is enabled, but cost data takes a few hours to start arriving and the previous month's backfill can take up to five days. You can keep going with the next steps meanwhile
4. In Roiva: Under Billing Export, enter the project and dataset the Billing export page shows, e.g. my-project.billing_export, and click Create Connection. Roiva finds the export table in that dataset (its full Table ID works too), creates a Google service account for this connection, and shows its email on the connection page
5. In Google Cloud: Grant that service account three read-only roles: Billing Account Viewer on the billing account, BigQuery Data Viewer on the export dataset, and BigQuery Job User on the dataset's project. The connection page lists the gcloud and bq commands, filled in, to paste into Cloud Shell or send to whoever administers your Google Cloud
6. Back in Roiva: The connection page checks each step as the service account and says what's still missing. Once it's all ticked, click Test Connection; your total Google Cloud spend syncs from there, with no budgets needed
7. In Google Cloud: To split spend by initiative (optional), create a budget for each one on the Budgets & alerts page (console.cloud.google.com/billing/budgets) whose name identifies it, e.g. roiva-invoice-ai, limited to the projects, services (such as Vertex AI or the Gemini API) or label that initiative uses. Creating budgets needs the Billing Account Administrator or Billing Account Costs Manager role
8. Back in Roiva: Link this connection to each initiative as a cost source and add a budget mapping with that initiative's budget name (or part of it)

Email hello@roiva.ai with what you're seeing, and you'll hear back within one business day.