OffDuty
The SRE you don't have — a reasoning layer that watches your infra, diagnoses incidents, and proposes fixes. Never executes.
OffDuty is a reasoning layer for small product teams — the 2–20 engineer companies where everyone has a day job and nobody’s on call at 3am. It watches your infrastructure, notices when something’s genuinely wrong, and then does the part that actually costs you sleep: the investigation. It reads the crashed container’s logs, checks what deployed an hour ago, forms a hypothesis, and hands you a root cause with a concrete fix to approve.
Nothing runs automatically. OffDuty diagnoses and proposes; you decide.
It’s deliberately not another dashboard. The missing piece usually isn’t more data — you already have metrics and logs, or a cluster small enough that you don’t. The missing piece is someone who looks at it when things break.
Open source, Apache-2.0. Early days — the name might not even survive.