The wall

90 minutes. 20 services. Two alerting policies.

17 of these p99-latency streams are healthy; 3 carry a real injected incident. Both policies watched every stream, minute by minute. Every stream is synthetic and labeled, reproducible from a frozen seed — and every alarm below was verified against the production engine.

POLICY A — STATIC MEAN+3σ THRESHOLD15 false pages

…in 90 minutes, on data where nothing was wrong. And it still caught only 2 of the 3 real incidents — one sailed through untouched.

Policy B — validanytime e-process0 false pages

…on the same 20 streams, and all 3 incidents caught — each page sent inside a stated false-alarm budget, with the evidence attached.

#100

#101

#102

#292incident

#103

#104

#105

#106

#107

#108

#109

#203incident

#110

#111

#112

#113

#210incident

#114

#115

#116

static mean+3σ page — threshold fit on each stream’s first 30 minutesvalidanytime page — false-alarm budget ≈ one per 2,000 minutes per streaminjected incident window (synthetic, labeled)

15 pages on nothing is how alert fatigue starts: the channel gets muted, and the incident that matters sails through. An alarm you can act on has to be rare on healthy data — that is the property the e-process guarantees, and the threshold rule can’t.

To be fair to the threshold: on the 2 incidents it did catch, it was faster — in minutes, +4 vs our +16, +4 vs our +17 (and it never caught the third; we paged that one +20 minutes after onset). The same sensitivity that catches in 4 minutes is what pages 15 times on nothing. We sell the quiet, not the sprint.

Replay your own history See how the guarantee works

Stop guessing whether it broke.

Point ValidAnytime at one stream and see whether — and on which day — it would have caught your last regression. On your own data, in minutes, free.

Start freeBacktest your own history

The backtest is free and runs in your browser — no signup, nothing uploaded.