SLA definitions
A service level agreement (SLA) is a written commitment about how quickly BuiltForProd acknowledges and works a request, measured in terms that both sides define the same way. This page defines those terms. The figures attached to them (how many minutes or hours, for which severity, during which hours) are not published here: they are stated in your customer agreement, and they apply to the Support SLA tier.
The terms
| Term | Meaning |
|---|---|
| Request | A single problem or question you raise through a support channel. One request has one severity and one owner at a time. |
| Severity | How much the problem hurts, assigned when the request is opened and revised if the facts change. Severity decides which response commitment applies. |
| Response time | The time from when you open a request to when a BuiltForProd engineer acknowledges it and starts working on it. It is not the time to a fix. |
| Resolution time | The time from when the request is opened to when the problem is fixed or a workaround restores service. Where an agreement states a resolution target, it is a target for the severity, not a guarantee that every root cause is removed within it. |
| Business hours | The hours during which response and resolution clocks run for the severities that are covered in business hours only. The agreement states the hours and the time zone. |
| Workaround | A change that restores service without removing the root cause; for example, promoting the previous release tag while the failing one is investigated. A workaround can close the resolution clock while a follow-up request tracks the root cause. |
| Hours | For On-Demand and Support SLA, the unit in which work is counted against the minimum (10 hours) or the pre-paid block (100 hours, expiring after one year). |
How severity is used
Severity is the language shared between your team's incident process and BuiltForProd's response commitment. The platform's own team workflow, delivered with the Baseline, uses three levels for the customer's internal process: production down, production degraded, and non-production. A support agreement uses its own severity scale, defined in the agreement; when you open a request, state the severity you believe applies and the facts behind it, and the engineer confirms it.
As a rule of thumb, "production down" means customers cannot use the service; "degraded" means they can, but something is wrong or slow; anything in sandbox, dev or staging is non-production. The tier you hold decides whether a response window is guaranteed.
What is not an SLA
- On-Demand support has no response commitment. It is pay-as-you-go help with a 10-hour minimum. The offerings note attaches SLAs to the Support SLA tier only.
- AWS service availability. The AWS Baseline runs on AWS services with their own service commitments; BuiltForProd's SLA covers BuiltForProd's response, not AWS uptime.
- Uptime of your workloads. BuiltForProd deploys and, under Managed, helps operate the platform; the availability your customers see depends on your application and your operations as well. The shared responsibility model draws the line.
When you are unsure which term applies, describe the impact in plain words when contacting support; the engineer maps it to the agreement.