Most dashboards stop at the chart. Something spikes, the line turns red, and you're left doing the actual investigation yourself — cross-referencing a deploy log, a Sentry issue, and a support ticket by hand, in three different tabs, while a customer waits.
We wanted the alert itself to do more of that work.
What opening an alert should give you
When you open something in the FounderOS Inbox, you don't just get a title and a timestamp. You get:
- An incident window chart — the metric in question, zoomed to the minutes around the spike, not a 30-day view that hides it.
- Ranked correlated events — deploys, errors, and other signals that happened in that same window, sorted by how likely each one is to be the cause.
- Evidence chips — the specific facts that led to that ranking, so the ranking is legible, not a black box.
- Actions — the concrete next step for this kind of alert, not a generic "view details" button.
That structure comes directly from how the original incident we built this around actually played out: an error spike on Trailwise, minutes after a deploy. The deploy wasn't the first thing anyone thought to check — it was the thing the ranking surfaced first, because it happened four minutes before the spike and touched the exact route that started failing.
Resolve and move on
Once you've dealt with it, Resolve moves the alert out of the open list and into a resolved-today section with a one-click Undo, in case you're wrong. No status meeting, no ticket to close in a separate tool. The Inbox is where triage happens, and it should let you leave it as fast as it let you understand the problem.
That's the bet behind FounderOS: the dashboard tells you something changed. The Inbox tells you why, and what to do about it.