MLA-C01 Deployment and Orchestration of ML Workflows Practice Question
A machine learning team has a model that needs to serve predictions with very low latency (under 10 ms) for a real-time web application. The model is a small ensemble of three neural networks that fits in memory. Which SageMaker inference option is MOST appropriate?
⚠ Common exam trap
Many candidates confuse 'low latency' with 'serverless' or 'asynchronous' options, not realizing that serverless inference has cold starts and asynchronous inference adds queueing delays, both of which break the sub-10 ms requirement.
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
✓
SageMaker real-time endpoint
SageMaker real-time endpoints are designed for low-latency, synchronous inference, making them the best fit for a model that must serve predictions in under 10 ms. Since the ensemble of three neural networks fits in memory, a real-time endpoint can keep the model loaded and respond to each request with minimal overhead, typically using HTTPS and the SageMaker InvokeEndpoint API.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
SageMaker batch transform
Why it's wrong here
Batch transform processes an entire dataset as an offline job, returning no synchronous per-request response, so sub-10 ms interactive serving is impossible. It tempts because it is the cheapest option for scoring large stored datasets, which is exactly when it would be correct.
- ✓
SageMaker real-time endpoint
Why this is correct
Real-time endpoints keep the model loaded on persistent instances and return predictions synchronously, avoiding the cold-start and queueing overhead of serverless inference. For a small in-memory ensemble needing sub-10 ms responses, this persistent hosting meets the latency requirement.
- ✗
SageMaker asynchronous inference
Why it's wrong here
Asynchronous inference queues requests and returns results via Amazon S3, so the caller cannot receive a prediction within 10 ms. It tempts because it handles large payloads and long processing times cost-effectively, which is when it would be the right choice.
- ✗
SageMaker serverless inference
Why it's wrong here
Serverless inference cold-starts and scales from zero, so it cannot guarantee sub-10 ms responses for a steady real-time workload. It tempts because it removes idle infrastructure cost for intermittent, spiky traffic, where occasional latency spikes are acceptable.
Go deeper
Related to this question
About these practice questions
One of 665 original MLA-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This MLA-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the MLA-C01 exam.