All videosOther topics0:36

Online evaluation: shadow, canary, A/B, and regression gates

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.

Links containing ?t= open the video at a specific second.

Video summary

The ideas to retain

01

Before exposing traffic, define the unit of change

A “candidate” should not simply mean “the new model.”

02

Shadow: observe the candidate before giving it authority over the response

Shadow traffic, also called mirroring, copies real requests to a candidate while the stable variant continues to serve the user.

03

Shadowing does not measure real user impact

The candidate response does not determine what the user sees. A shadow therefore cannot directly observe:

Key moments

Jump directly to a section

  1. Shadow observes traffic without changing the decision
  2. Canary increases exposure only with green guardrails
  3. A/B measures effect; the regression gate promotes or rolls back