ValidAnytime
  • Try it free
  • Observatory
  • Pricing
  • Docs
Sign inStart free
ValidAnytime

A guaranteed false-alarm budget across your whole fleet, valid no matter how often you look.

Made by Compiled Intelligence — a frontier AI lab working on quantitative finance from first principles; ValidAnytime is the monitoring we built for our own model fleets, productized.

Product

  • How it works
  • Pricing
  • Observatory
  • Try it free
  • Dashboard

Solutions

  • Who it's for
  • For LLM engineers
  • For MLOps engineers
  • For SRE & platform
  • LLM eval regression
  • Model drift detection
  • Alert fatigue

Developers

  • Docs
  • Quickstart
  • Python SDK
  • API reference

Learn

  • Benchmark
  • Detector guides
  • Glossary
  • Compare
  • Why dashboards lie
  • Blog
  • Anytime-valid 101
  • Winter Storm Uri case study
  • Incident ledger
  • Observatory methodology

Company

  • Contact
  • Privacy
  • Terms

Observatory alarms & product updates, by email

Rare emails only. We store your address and the tag “footer” — nothing else. Unsubscribe anytime.

© 2026 ValidAnytime. All rights reserved.

Every alarm ships with its guarantee_tag and theorem_ref.

Glossary

The peeking problem

The peeking problem is the reason a metric can look 'significant' just because you checked it too many times.

Also known as: optional stopping / repeated significance testing

The peeking problem is the trap that catches almost everyone: you have a test, you check it repeatedly as data comes in, and you stop the moment it looks bad. Even if nothing is actually wrong, keep looking long enough and pure noise will eventually cross any fixed line. Every extra look is another roll of the dice, and classical tests were never designed to be checked more than once.

It is exactly why monitoring dashboards cry wolf. A threshold that would false-alarm 5% of the time if you looked once will false-alarm far more often if you look every minute. Teams respond by muting alerts — and then miss the real regression.

Anytime-valid methods solve the peeking problem at the root: their guarantees are proven to hold across all looks at once, so continuous checking is not just allowed, it is the intended use.

Go deeper

  • See it happen: the peeking demo
  • Why your dashboard is lying to you
  • The fix for alert fatigue
  • For LLM engineers
  • How much classical detectors over-alarm — the benchmark

Related terms

  • Anytime-valid inferenceAnytime-valid inference is a way of testing that stays statistically valid no matter how often you look at the results.
  • Sequential testingSequential testing is the practice of testing a hypothesis as data arrives, deciding to stop as soon as the evidence is conclusive.
  • E-processAn e-process is a running score of evidence against 'nothing has changed'; its value at any moment is an e-value, and it stays valid at every look.
  • Always-valid p-valueAn always-valid p-value is a p-value you are allowed to read at any moment — it never gets less trustworthy the more often you check.
  • False discovery rateThe false discovery rate is the fraction of fired alarms that turn out to be false — the number you actually want to control across a fleet.
  • Changepoint detectionChangepoint detection is the task of spotting the moment a metric's behavior genuinely shifts — as opposed to normal noise wobbling around.
  • SPC for MLSPC for ML is the practice of putting machine-learning metrics under statistical process control — control charts with explicit rules, not eyeballed dashboards.

Put the theory to work.

ValidAnytime turns these ideas into a live alarm you can trust — valid no matter how often you look. Prove it on your own data, free.

Start freeTry the detector