Zeno Check
Should I trust this forecast?
Zeno checks predictions from models you already use. See where they may fail, the adjusted forecast, and what happened.
Ensemble weather
Sydney temperature · 24 hours ahead
Zeno moved the panel estimate toward the outcome and flagged two narrow forecasts. UKMO missed; ECMWF's range held, so that one was a false alarm.
Served answer
Zeno correction
26.5°CZeno
25.1°Cactual 22.2°C
panel
26.5°C
Zeno
25.1°C
actual
22.2°C
% = chance that forecaster misses its own 90% range · not a chance of “surprise” or profit
01 · review budget
How many alerts are worth it?
Across the later test weeks, here is what acting on Zeno's alerts would have meant.
of 100 alerts
62 real misses · 38 false alarms
share of all confident misses caught
33%
The rest still slip through. Alerts fire above the level that flagged the riskiest 10% of training cases.
02 · weather pilot
How a weather team uses it.
The first pilot is weather operations. Energy follows once operational decisions are tested too.
01
Connect the forecast feeds you already receive
02
Pick location, variable and deadline
03
Get the adjusted forecast, failure risks and validation scope
04
Review the flagged forecasts
05
Compare with observed outcomes
03 · source status
Correction must be earned.
Each source is measured on later data. Where Zeno did not improve, the reference is kept.
Crypto price ranges
Zeno correction
+11.7%
Ensemble weather
Zeno correction
+5.6%
Deterministic weather
correction · no history
+6.2%
COVID-19 hospitalisations
correction · no history
+11.8%
Sea conditions
correction · no history
+3.2%
Prediction markets
reference kept
RSV
reference kept
FluSight
reference kept
ForecastBench
reference kept
Numinous AI forecasters
learn-only
04 · for developers
Need the machinery?
The public interface stays simple. Model files and integrations remain available separately.