TV Ops Dashboard

The story

I wanted a single screen in my office that answered: "Is anything on fire?" and "What needs attention today?" The problem: most dashboards are either too generic, too noisy, or built for one tool at a time. I wanted something opinionated, high-signal, and easy to extend — a simple rotation of views I can tweak as my stack (and attention span) changes. So I built a TV-friendly ops dashboard that sits up all day and cycles through pipelines, jobs, failures, and other "don't ignore this" signals.

Highlights

• Rotating view system with configurable timing/weighting (so the important stuff shows up more).
• High-signal 'failures' view: what's broken, how stale it is, and where to click to fix it.
• Kiosk-friendly UX: readable from across the room, minimal interaction required.
• Acknowledgements/snoozes so one flaky job doesn't hijack the whole screen forever.
• Built to be extended: add a new view, wire a data source, ship it.

What it shows

• CI/CD pipelines and job failures
• 'Stuck' or stale runs
• The handful of operational metrics I actually react to
• Anything else that earns a spot on the wall

Tech stack

Lessons learned

• If it's not visible, it doesn't exist (until it breaks at the worst possible time).
• High-signal beats high-volume. A wall of charts is just expensive wallpaper.
• 'Configurable' is only useful if it's easy. If it needs a redeploy, you won't use it.