Note /
Our SLA, read honestly
Support promises are easy to write and hard to read. Here is ours, taken apart line by line, including the parts that limit it.
Response is guaranteed, resolution is best-effort
These are different promises and we keep them separate on purpose. You will hear back within 24 hours, guaranteed. Whether the issue can be resolved, and how fast, depends on what it is: a configuration fault in the deployment we installed lands squarely in scope, while a model misbehaving or a chat platform outage is not something anyone can honestly guarantee away.
What 24/7 agent-assisted means
Support agents answer around the clock, which is how a small firm offers a 24-hour guaranteed response without pretending to staff a follow-the-sun help desk. When an issue needs a person, it escalates to a human the next business day. We run these support agents on our own deployment, the same way yours will run.
What is in scope
Any issue with the deployment as we installed and configured it, running the pinned stable version. Pinning matters: updates reach you only after they pass our staging, so the version you run is one we have already tested.
What is not, and why we publish it
Upgrades you start yourself, third-party outages, LLM provider costs and model behaviour, updates applied before staging sign-off, data loss where the backup add-on was declined, consequences of prompt injection or social engineering after the posture was signed off, and new features or integrations, which are sold as add-ons rather than support. Every exclusion is on the pricing page in plain type, because an SLA you cannot read the edges of is not a promise, it is a surprise scheduled for later.