MLA-C01 Question #138: Real Exam Question with Answer & Explanation

Question

An ML engineer needs to deploy four ML models in an Amazon SageMaker inference pipeline. The models were built with different frameworks. The ML engineer also needs to give clients the ability to use the invoke_endpoint call to perform inference for each model. Which solution will meet these requirements MOST cost-effectively?

Options

• ACreate a SageMaker multi-model endpoint.
• BCreate a SageMaker multi-container endpoint.
• CCreate multiple SageMaker single-model endpoints.
• DRun a SparkML job to generate multiple endpoints.

Explanation

A SageMaker multi-container endpoint hosts multiple containers - each with its own framework - on a single endpoint, and clients can target individual containers via invoke_endpoint using the TargetContainerHostname parameter. This satisfies both the multi-framework requirement and the per-model invocation requirement while sharing underlying compute, making it the most cost-effective fit.

Why the distractors fail:

• A (Multi-model endpoint): Designed for many models that share the same serving container/framework - it cannot support different frameworks per model.
• C (Multiple single-model endpoints): Technically works but spins up four separate instances, multiplying infrastructure cost unnecessarily.
• D (SparkML job): SparkML is a batch data-processing tool, not an inference hosting solution - it does not create invocable endpoints.

Memory tip: Think of it as an apartment building vs. separate houses. A multi-container endpoint is one building with different-styled units (frameworks) under one roof (one endpoint cost). A multi-model endpoint is one building where every unit must look identical (same framework). Multiple single-model endpoints are four separate houses - functional, but expensive.

No community discussion yet for this question.