---
title: 'Online evaluation: shadow, canary, A/B, and regression gates'
seo_title: 'Online evaluation: shadow, canary, A/B, and regression gates — video'
description: 'Online evaluation increases exposure under control: shadow observes without deciding, canary limits blast radius, A/B estimates effect, and guardrails or regression gates stop an unsafe promotion.'
keywords: online evaluation, shadow traffic, canary, A/B testing, guardrails, regression gates, AI systems, progressive delivery
date: '2026-09-13T00:00:00+00:00'
robots: index,follow,max-snippet:-1,max-image-preview:large,max-video-preview:-1
hide:
- toc
- navigation
video_watch_page: true
---


<div class="s5-video-watch" data-s5-video-watch data-video-id="videos-series-evaluating-ai-systems-production-05-online-evaluation-shadow-canary-ab-guardrails-regression-gates-md">
  <header class="s5-video-watch__header">
    <div class="s5-video-watch__crumbs"><a href="https://5sigmas.com/en/videos/">All videos</a><span>Other topics</span><span>0:36</span></div>
    <h1>Online evaluation: shadow, canary, A/B, and regression gates</h1><p>Online evaluation increases exposure under control: shadow observes without deciding, canary limits blast radius, A/B estimates effect, and guardrails or regression gates stop an unsafe promotion.</p>
  </header>
  <div class="s5-video-watch__player">
    <video controls crossorigin="anonymous" preload="metadata" poster="/en/series/evaluating-ai-systems-production/05-online-evaluation-shadow-canary-ab-guardrails-regression-gates.jpg" playsinline data-s5-watch-player><source src="/en/series/evaluating-ai-systems-production/05-online-evaluation-shadow-canary-ab-guardrails-regression-gates.mp4" type="video/mp4">Your browser does not support the video element.</video>
    <p>Links containing <code>?t=</code> open the video at a specific second.</p>
  </div>
  <section class="s5-video-watch__summary" aria-labelledby="video-summary-title">
    <div class="s5-video-watch__section-head"><span class="s5-eyebrow">Video summary</span><h2 id="video-summary-title">The ideas to retain</h2></div>
    <div class="s5-video-watch__snippet-grid"><article><span>01</span><h2>Before exposing traffic, define the unit of change</h2><p>A “candidate” should not simply mean “the new model.”</p></article>
<article><span>02</span><h2>Shadow: observe the candidate before giving it authority over the response</h2><p>Shadow traffic, also called mirroring, copies real requests to a candidate while the stable variant continues to serve the user.</p></article>
<article><span>03</span><h2>Shadowing does not measure real user impact</h2><p>The candidate response does not determine what the user sees. A shadow therefore cannot directly observe:</p></article></div>
  </section>
  <section class="s5-video-watch__chapters" aria-labelledby="video-chapters-title"><div><span class="s5-eyebrow">Key moments</span><h2 id="video-chapters-title">Jump directly to a section</h2></div><ol><li><a href="?t=0" data-s5-video-seek="0"><time>0:00</time><span>Shadow observes traffic without changing the decision</span></a></li>
<li><a href="?t=12" data-s5-video-seek="12"><time>0:12</time><span>Canary increases exposure only with green guardrails</span></a></li>
<li><a href="?t=24" data-s5-video-seek="24"><time>0:24</time><span>A/B measures effect; the regression gate promotes or rolls back</span></a></li></ol></section>
  
  <aside class="s5-video-watch__source"><div><span class="s5-eyebrow">Context and evidence</span><h2>Continue with the full article</h2><p>The chapter develops the mechanism, primary sources, limitations and connections to the rest of the series.</p></div><a class="s5-video-watch__source-link" href="https://5sigmas.com/en/series/evaluating-ai-systems-production/05-online-evaluation-shadow-canary-ab-guardrails-regression-gates/">Read the article →</a></aside>
  <section class="s5-video-watch__related" aria-labelledby="related-videos-title"><div class="s5-video-watch__section-head"><span class="s5-eyebrow">Next step</span><h2 id="related-videos-title">Related videos</h2></div><div class="s5-video-watch__related-grid"><article><a href="https://5sigmas.com/en/videos/series/evaluating-ai-systems-production/06-observability-failure-taxonomies-production-eval-repair-feedback-loops/"><img src="https://5sigmas.com/en/series/evaluating-ai-systems-production/06-observability-failure-taxonomies-production-eval-repair-feedback-loops.jpg" alt="" loading="lazy" width="1280" height="720"><span>Other topics · 0:36</span><strong>Observability, failure taxonomies, and feedback loops</strong></a></article>
<article><a href="https://5sigmas.com/en/videos/series/evaluating-ai-systems-production/04-evaluacion-trayectorias-agentes-tools-exito-eficiencia-recuperacion-policy/"><img src="https://5sigmas.com/en/series/evaluating-ai-systems-production/04-evaluacion-trayectorias-agentes-tools-exito-eficiencia-recuperacion-policy.jpg" alt="" loading="lazy" width="1280" height="720"><span>Other topics · 0:36</span><strong>Agent trajectories: success, efficiency, recovery, and policy</strong></a></article>
<article><a href="https://5sigmas.com/en/videos/series/llm-inference-engineering-economics/06-benchmarking-inference-cost-task-throughput-latency-energy-hardware-constraints/"><img src="https://5sigmas.com/en/series/llm-inference-engineering-economics/06-benchmarking-inference-cost-task-throughput-latency-energy-hardware-constraints.jpg" alt="" loading="lazy" width="1280" height="720"><span>Other topics · 0:36</span><strong>Benchmarking: cost, throughput, latency, and energy</strong></a></article></div></section>
</div>