Cost as Code
Cloud and AI spend, managed the way good teams manage code.
In 30 seconds
Cost as Code is managing cloud and AI spend the way software teams manage code.
- Every saving is a change you review and approve.
- Every spend rule lives in a file in your repo.
- Every result is measured on the bill.
Every saving goes through the same loop: find it, fix it with a change you approve, prove it on the bill, and prevent it from coming back. The lifecycle explains each step, and the principles say why the method works this way.
Two modules, one loop
Available now
Savings as Code
Every saving is a pull request you approve, then measured on the bill.
Coming soon
Spend as Code
Your budgets and limits for cloud and AI in one file, checked wherever money is spent.
Savings as Code removes the waste. Spend as Code keeps it from coming back. For AI, Savings as Code starts with right-modeling: the smallest model that meets the quality bar. In practice shows both on AWS, with your own tools or with our service.
Cost as Code, FinOps as Code and policy as code
| Term | What it means |
|---|---|
| FinOps as Code | A related term used in several ways, from configuring cost tools in code to keeping cost rules, budgets and policies in version-controlled files. More. |
| Policy as code | Rules are written in code and checked by tools, often for security or compliance. |
| Cost as Code | Every saving is a reviewed change measured on the bill, and the spend rules for cloud and AI live in one file. |
Where the method comes from, and why it is called "as code": origins and why "as code". The words used on this site: glossary.
Who writes this
Cost as Code is written by MTO FinOps, the cost line of MoreThanOps. We run Savings as Code as a service on AWS today. About us.
Latest from the blog
- Implement AWS cost recommendations as pull requests
AWS cost recommendations stall between the finding and the change. Turning each saving into a pull request you approve, then measuring it, fixes that.
- Five LLM cost levers to check before you switch providers
Before moving to a cheaper AI provider, check five levers that reduce LLM costs where you are: model routing, caching, batch, output limits and context.