Operations, without the noise

Read the whole week before your coffee cools.

Graphite panels, dense tables, honest charts. Nightdesk puts the numbers your team actually argues about on one screen — and holds the color back for things that moved.

Book a walkthrough
SOC 2 Type II · 99.99% uptime · Read-only by default
Signal · this week+18.4%
412
▲ 64 resolved without escalation
Noise · suppressed38 alerts
7
▼ 82% fewer pages than last week
nightdesk.app / operations / week-42
⌘K◐
Operations overviewWeek 42 · 09:41 UTC
Last 7 daysAll servicesCompare: last weekExport
Requests18.4M▲ 6.2%
p99 latency212ms▲ 4.1%
Error rate0.07%▼ 0.02%
Spend$12,480▲ 2.8%
Throughput vs. latency● Requests● p99
ServiceOwnerUptimep99Trend
ledger-apiPayments99.997%182ms▲ stable
auth-edgePlatform99.991%94ms▲ stable
search-indexData99.982%240ms▼ degraded
billing-workerPayments99.999%110ms▲ stable
notify-fanoutGrowth99.994%76ms▲ stable
Made for the on-call brain

Dense where it matters. Quiet everywhere else.

The screen should tell you what changed since you last looked. Everything else is archive.

▤

Dense tables, honest columns

Every service, owner, p99 and trend in one scan. Sort, pin and share a view without exporting to a spreadsheet.

◫

Timelines that line up

Deploys, incidents and alerts share one time axis. The cause is usually the thing three rows above the symptom.

◔

Color held in reserve

Gray is the default. Lime means gain, amber means drift, red means wake someone up. Nothing else gets a color.

““We stopped paging people for things that were already recovering. Nightdesk made the difference between an alert and a decision visible.””
Mara Velasquez · Head of Platform, Fieldwork

Put the whole week on one screen.

Connect your first service in ten minutes. Read-only until you say otherwise.