---
title: "A/B and shadow deployment | Skills We Assign For | Azendo"
description: "Shadow and A/B model deployment — validating on real traffic safely, plus the Azendo roles assigned for it."
url: "https://azendo.co/skills/a-b-and-shadow-deployment/"
---

[Skills](https://azendo.co/skills/) MLOps and model delivery 

# A/B and shadow deployment.

Shadow deployment runs a new model alongside the current one on real traffic without serving its predictions. A/B deployment serves a fraction of traffic from the new model and compares outcomes.

## Where A/B and shadow deployment fits on a long engagement.

Shadow deployment is the safest way to validate a model on production traffic. The new model sees real inputs and its predictions are recorded and compared, but nothing it says reaches a user — so a model that behaves badly on live data is discovered without anyone being affected.

It answers questions offline evaluation cannot. Real traffic contains distributions, edge cases and malformed inputs that a curated test set does not, and latency under production load is only measurable under production load.

## What an assigned team does with A/B and shadow deployment.

A/B deployment is the next step and needs the same discipline as any experiment: a predefined metric, a sample size, and a decision rule agreed before results are visible.

Holding that sequence is what makes the comparison trustworthy, and it is the standard applied under [ai engineering services](https://azendo.co/services/ai-engineering/).

## What we use A/B and shadow deployment for.

* Validation on real traffic with no risk Predictions recorded rather than served, so bad behaviour affects nobody.
* Latency measured under real load Production performance established before the model serves anyone.
* A/B comparison with a stated rule Metric and decision threshold agreed before the results are known.

## How A/B and shadow deployment capacity is assigned.

Deployment strategy is assigned inside AI capacity, with shadow validation preferred before any model serves live predictions.

## Roles we assign A/B and shadow deployment for

* [Model Deployment Engineer MLOps engineering](https://azendo.co/services/mlops-engineering/model-deployment-engineer/)

## Service lines it sits in

* [MLOps engineering](https://azendo.co/services/mlops-engineering/)

Capacity is agreed as a committed monthly capacity across a discipline, not per skill.

## Related in mlops and model delivery

* [MLOps handoff — skill we assign for](https://azendo.co/skills/mlops-handoff/)
* [ONNX — skill we assign for](https://azendo.co/skills/onnx/)
* [TensorRT — skill we assign for](https://azendo.co/skills/tensorrt/)
* [Model registry — skill we assign for](https://azendo.co/skills/model-registry/)
* [Weights & Biases — skill we assign for](https://azendo.co/skills/weights-and-biases/)
* [Comet — skill we assign for](https://azendo.co/skills/comet/)
* [Neptune.ai — skill we assign for](https://azendo.co/skills/neptune-ai/)
* [DVC — skill we assign for](https://azendo.co/skills/dvc/)

## Tell us what your roadmap needs A/B and shadow deployment for.

A service delivery manager replies with the disciplines we would assign, the monthly capacity and what the first month looks like.

[All skills we assign for](https://azendo.co/skills/)