Live demo
14 series in. 4 alerts out.
An inference fleet of forty H100 pools, 60 days, 14 series. On day 43 a driver rollout reached one pool and stopped there. Uptime stayed at 99.94%, request volume kept its weekly shape, and the fleet token rate never moved. Find the pool.
Every number on this page is computed in your browser right now, by the same CUSUM and Bayesian changepoint code that runs in production. Nothing here is a screenshot.
The comparison that matters
Against the rule most teams actually ship
The standard configuration is a percentage or sigma threshold against a trailing mean, evaluated per metric. Run it over the exact same 14 series and count what lands in the channel.
The false-alarm count is not a judgement call. The incident was injected at a known index, so every naive alert before that index is wrong by construction.
What the engine found
Detection run
| Metric | Trend | Baseline | Now | Change | Confidence | Method |
|---|---|---|---|---|---|---|
Thermal throttle events pool=a3-h100-04 | 35.07 | 183.94 | +425% | 99.9% | CUSUM | |
Cost per million tokens pool=a3-h100-04 | $2.84 | $3.88 | +36.9% | 99.9% | CUSUM | |
Tokens per second pool=a3-h100-04 | 4187.483 | 3023.752 | −27.8% | 99.9% | CUSUM | |
HBM bandwidth utilisation pool=a3-h100-04 | 71.2% | 51.6% | −27.5% | 99.9% | CUSUM |
Show your work
Thermal throttle events · pool=a3-h100-04
Baseline mean μ = 0.0024 and σ = 0.1761, both computed from the first two thirds of the day-over-day percentage changes. The slack k = 0.0881 is half a sigma, and the decision boundary h = 0.7046 is four. The accumulator runs on percentage changes rather than raw levels so a healthy growing series cannot drift across the boundary on its own.
alert if |Δ| > 2.5σ vs 7-day mean fired 3 times across this one metric, including 2 before anything was actually wrong. The statistical pipeline sent one message, on the day the regime actually changed.
What lands in Slack
One message, with the work already done
- 1.Roll a3-h100-04 back to driver 550.90 and hold it for 48 hours.
- 2.Compare per-node throttle counts inside the pool to separate a fleet-level driver issue from two bad cards.
- 3.Check the rack inlet temperature log for Aug 11 before blaming the driver.
The written cause is generated only after the statistics confirm the change. The model never decides whether something is an anomaly — it explains one that has already been established. Getting that order backwards is how these products hallucinate.