PMLE Automating and Orchestrating ML Pipelines Practice Question

A company has a CI/CD pipeline that retrains a model every time new training data is available. They want to automatically deploy the new model to production only if it passes a set of evaluation tests on a staging environment. Which approach best implements this?

⚠ Common exam trap

PMLE often tests whether candidates confuse 'automated deployment' with 'automated promotion' — the trap is choosing an option that deploys automatically but omits the evaluation gate, or one that evaluates but requires manual promotion.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Implement a two-stage pipeline: train and deploy to staging, run evaluation tests, and if passed, deploy to production using conditional logic.

The correct approach is a two-stage pipeline: train and deploy to a staging environment, run automated evaluation tests against the staged model, and use conditional logic to promote to production only if the tests pass. This implements a proper ML CI/CD gate that prevents regressions from reaching production.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Implement a two-stage pipeline: train and deploy to staging, run evaluation tests, and if passed, deploy to production using conditional logic.
Why this is correct
A two-stage pipeline trains and deploys to staging, runs evaluation tests, then uses conditional logic to promote to production only on passing results. This gates deployment on measured quality, satisfying the requirement to release solely when evaluation succeeds.
✗
Use Cloud Build to trigger a training job and then a separate deployment job without evaluation.
Why it's wrong here
Chaining training to deployment without an evaluation step means the pipeline promotes every model unconditionally, ignoring the staging test requirement. It is tempting because Cloud Build orchestrates multi-step CI/CD jobs well when the only goal is automated delivery with no quality gate.
✗
Use a single Vertex AI pipeline that trains and deploys to staging, then manually promote.
Why it's wrong here
A single pipeline ending in manual promotion replaces the required automatic deployment with human intervention, so passing models are not released automatically. It is tempting because one Vertex AI pipeline covering train-and-stage is the standard pattern when a human must approve each release.
✗
Train and deploy directly to production in one pipeline.
Why it's wrong here
Deploying straight to production skips the staging evaluation gate the stem requires, so a failing model reaches users unchecked. It is tempting because a single train-and-deploy pipeline is the standard pattern when no quality gate exists and every trained model is trusted to ship immediately.

Go deeper

PMLE Automating and Orchestrating ML Pipelines Practice Question

Why each option matters

Option-by-option breakdown

Related to this question