Verified savings, in production.
Three deployments, three industries, one audit row — the same one-PR, one-deploy-timestamp, one-measurement standard /methodology and /security publish. Pinegrove Co., Helio Trust, Northwind Labs.
Three deployments. One audit standard.
Each saving carries one PR, one deploy timestamp, and one post-deploy measurement window — the same audit row /methodology publishes. If a recommendation does not survive the measurement window the credit is reversed.
AWS · Terraform · OIDC
Annual verified savings
From read-only connection to first signed diff
Savings measured per /methodology PILLARS[0] — one post-deploy window per PR; credit reversed if the window fails.
Pinegrove inherited a heavy dev-test footprint after its Series B. Finance couldn't reconcile rightsizing numbers and the CFO refused to approve changes without a paper trail. Tallywyrm connected through read-only OIDC credentials, published vendor-signed Terraform PRs one diff at a time, and attached one deploy timestamp and one measurement to every change. Ninety days in: $1.4M of annual verified savings, 412 idle instances shut down, no write surface against the Pinegrove account.
GCP · Terraform
Annual verified savings
From read-only connection to first signed diff
Savings methodology: /methodology PILLARS[0] — each Compute Engine and Cloud SQL recommendation carries one signed diff and one post-deploy measurement.
Helio Trust's security-review cycle meant every infrastructure change needed a signed diff before it could merge. The team had no baseline for over-provisioned Compute Engine and Cloud SQL resources. Tallywyrm scanned GCP inventory under read-only credentials and posted its first signed Terraform PR in under two hours. Each recommendation landed with one deploy timestamp and one measurement window, giving the security team an audit row they could present to customers without extra documentation.
AWS + GCP
Annual verified savings
From read-only connection to first signed diff
Full event log available at /dashboard/audit-trail (sign-in required) — every idle EMR cluster and BigQuery slot reservation carries its own audit row.
Northwind Labs ran a mixed AWS and GCP footprint and needed a unified spend view before its IPO audit. Two clouds, two billing APIs, one paper trail. Tallywyrm scanned inventory across both providers under read-only credentials, surfaced idle EMR clusters and oversized BigQuery slot reservations, and posted vendor-signed PRs the data platform team could review in one queue. The CFO and auditors verified every line directly from the audit-trail dashboard. Total reconciled savings: $920K / yr.
The same audit, applied to your invoice.
Book a working demo and a FinOps analyst will walk through the same one-PR, one-deploy-timestamp, one-measurement playbook that produced the three case studies above — using your cloud account, your Terraform state, your numbers.