AI-102 Plan and manage an Azure AI solution Practice Question
Your organization is migrating on-premises machine learning models to Azure. The models are used for real-time inference. You need to choose a service that provides managed endpoints with autoscaling and supports custom containers. Which service should you use?
⚠ Common exam trap
A common mix-up: candidates confuse Azure Kubernetes Service (AKS) as the only option for custom containers, overlooking that Azure Machine Learning managed endpoints natively support custom containers with autoscaling, eliminating the operational burden of managing a Kubernetes cluster.
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
✓
Azure Machine Learning managed online endpoints
Azure Machine Learning managed online endpoints are the correct choice because they provide fully managed, autoscaling endpoints specifically designed for real-time inference. They support custom container images, allowing you to deploy any model packaged as a Docker container, and handle traffic splitting, health checks, and scaling automatically without managing underlying infrastructure.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Azure Machine Learning managed online endpoints
Why this is correct
Azure Machine Learning managed online endpoints satisfy the real-time inference constraint by providing autoscaling managed endpoints that deploy custom containers. Unlike batch endpoints, which process asynchronous jobs, managed online endpoints expose a REST URI for low-latency scoring, and Microsoft Entra ID handles authentication, meeting the migration requirement without managing infrastructure.
- ✗
Azure Functions
Why it's wrong here
Azure Functions runs event-driven code with consumption-based triggers and bindings; it provides no managed model endpoints, no model registry integration and no container-based scoring server for real-time inference. It is tempting because it scales automatically and can host containers, and would be correct for lightweight event processing around a model rather than serving it.
- ✗
Azure Kubernetes Service (AKS) with manual scaling
Why it's wrong here
Manual scaling cannot satisfy the autoscaling requirement, and AKS makes you manage node pools, ingress and endpoint lifecycle yourself rather than consuming managed endpoints. It is tempting because AKS hosts custom containers for real-time inference, and would be right when you need full Kubernetes control over networking, GPUs or existing cluster investments.
- ✗
Azure AI Services custom vision
Why it's wrong here
Custom Vision trains and serves only its own image classification and object detection models; it cannot host arbitrary custom containers or expose managed inference endpoints for your own model artefacts. It is tempting because it offers managed endpoints with autoscaling, and would be correct for image labelling scenarios where a prebuilt vision model suffices.
Go deeper
Related to this question
About these practice questions
Courseiva writes every AI-102 question from scratch — 761 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI-102 practice question is part of Courseiva's free Microsoft 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 AI-102 exam.