Yearly running cost Assuming 2,400 changes a year
By hand $28.8K / year 480 hours of work
Subscription $12K+ / year
This workflow $62 / year Machine usage only; setup, hosting and review are extra. How should it receive your infrastructure changes? Compare before-and-after costs, check your budget rules and show what is not priced.
One change includes up to 100 resource entries and 20 cost policy checks.
Upload infrastructure snapshots Review a proposed change before approval $720 – 2,160 setup Connect my repository Review changes in your existing process $1,080 – 3,240 setup
Which repository should it connect to?GitHub GitLab Azure DevOps My repository system Your engineers approve changes. No automatic merges, infrastructure changes, resource deletion or cloud commitment purchases.
What is included Compare costs and apply your approved rules. Return resource-level evidence, assumptions and anything that could not be priced.
Start with sanitised infrastructure or cost snapshots. Add a repository connection for repeat reviews. The starting scope covers agreed Terraform resources on your chosen cloud, not every service or cloud.
Cloud providers in the starting scope: up to 1 provider.
Repositories in the starting scope: up to 1 repository.
Cost calculation tolerance: up to 1 cent.
How reliable should the cost and policy checks be? A useful review calculates supported costs correctly, catches your rules and never hides unknown charges.
Standard Verify the cost comparison and common budget rules, with unknown costs sent for review. Correct cost calculations: ≥100% Cost policy breaches found: ≥95% Included Strict Add independent verification and more difficult pricing and rule cases. Correct cost calculations: ≥100% Cost policy breaches found: ≥98% Setup +$144 – 480 My own targets Agree your own cloud services, cost policies and acceptance requirements. Quote separately
How these standards are measured Test independently calculated prices and labelled policy cases. Check cost arithmetic, missed rules, false flags and evidence. Unknown usage, unsupported resources and stale prices must remain visible.
Targets for your selected standard What is checked Target Correct cost calculationsCount correct price, quantity and total calculations against independently computed references. Fix the region, currency, price date, usage and billing convention. Missing or unsupported entries cannot pass as priced calculations. ≥100% Cost policy breaches foundCorrectly flagged breaches divided by all labelled breaches of the agreed policies. Missed and deferred breaches remain in the denominator. ≥95% Correct policy flagsConfirmed breaches divided by all proposed policy flags. A missing usage estimate is uncertainty, not proof of a budget violation. ≥95% Valid cost evidenceEach finding identifies the resource, input version, price source and date, region, currency and usage assumption. Unsupported explanations fail. ≥100% Unknown costs sent for reviewAll missing usage, unsupported resources and stale price cases remain visible and go to engineering review. A partial estimate must not be described as the complete bill. ≥100%
Use deterministic price calculations. Estimated costs are not the actual cloud bill. Unknown usage or unsupported resources must not become zero. Engineers retain approval and deployment authority.
How quickly do you need each change reviewed? Choose how quickly the cost comparison and findings should be ready.
Within 2 minutes Included Within 60 seconds Setup +$144 – 480 Within 30 seconds Setup +$288 – 840
Timing details Measure from complete approved inputs to the stored review, including price lookup, queues and retries. Creating a Terraform plan or waiting for engineering approval is separate.
The target applies to at least 95% of agreed test runs, with 2 in progress at a time.
How much do you want to spend per change? Choose the machine budget for calculating costs and explaining the findings.
Up to $0.25 Included Up to $0.15 Setup +$72 – 360 Up to $0.10 Setup +$216 – 600
Cost details Includes model calls, compute, retries and shared hosting at the stated workload. Staff review and software subscriptions are separate. Exceeding a price-service allowance needs repricing.
Reference machine cost: $0.13 – 0.23 per change at 200 changes a month. The selected cap is a target to test, not a replacement for this estimate.
Where do you want it to run? Run in your cloud or on your own machine. Choose whether approved cost data can go to external services.
Run it onYour cloud Local environment AI model accessApproved model API No external calls
Data and access details Runs in an account you control, with restricted access and a review log.
Use approved price lookups and model calls on sanitised cost evidence. Agree which data may leave your environment.
Sanitise inputs before processing. Private mode uses local files, a dated approved price table and private inference; it cannot retrieve live prices or read a remote repository.
How do you want to use it? Use a review web page, your existing AI agent or a dedicated cost-review agent.
A review web page Upload cost snapshots and review the change, assumptions and policy findings. Included My existing AI agent Ask your agent to review an infrastructure change and retrieve its cost evidence. Setup +$72 – 240 A dedicated agent Use a separate cost-review agent to explain proposed changes. Setup +$144 – 480