AI ART Discover · Learn · Create
S AI Glossary Term • Term 192 of 224
Plain English Definition

What does “Shadow Deployment” mean?

Core Definition

Running a model in parallel with production traffic without affecting user-facing decisions.

Operational Mechanism

How Shadow Deployment Works

Shadow Deployment: Live traffic feed into Both models receive the same requests, which produces Current model: answers users, New model: logged only.

Real-World Scenario

Practical Example

A bank runs a new fraud model alongside the current one on real transactions. The new model's decisions are recorded but not acted on, and the team compares them before switching.

Core Significance

Why It Matters

It tests a model on real traffic without exposing users to its mistakes. It costs extra compute while running.

Expert Q&A

Common Questions About Shadow Deployment

Q: Shadow vs A/B test?

In a shadow test, the new model's output is not shown to users. In an A/B test, some users receive it.

Q: What does it reveal?

How the new model behaves on real inputs, including rare cases that test data lacks.

Q: What are its limits?

Users do not react to the new output, so effects on behavior cannot be measured.