AI0-001 AI Infrastructure and Technologies Practice Question
A media company trains a video tagging model on a large dataset in the cloud. The model will run inference on-premises in a facility with intermittent network connectivity, and the operations team wants to avoid re-authoring the model for each target runtime. Which deployment artifact best meets these constraints?
⚠ Common exam trap
The trap here is equating containerization with runtime portability, when a container still embeds one specific framework build and does not standardize the model graph.
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
✓
An ONNX model executed with the ONNX Runtime
ONNX defines a standardized computation graph, and ONNX Runtime executes that graph across platforms and hardware without requiring the original training framework. Exporting to ONNX gives the media company a single artifact that runs on-premises in a disconnected facility and removes the need to re-author the model for each inference runtime, satisfying both stated constraints.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
A pickle file of the scikit-learn estimator
Why it's wrong here
Pickle serializes Python objects and depends on the exact library versions and class definitions used at training time. It is not a portable graph format, it is unsafe to load from untrusted sources, and it cannot represent the deep learning video tagging model described here. This choice fails both the portability and the framework-independence requirements.
- ✓
An ONNX model executed with the ONNX Runtime
Why this is correct
ONNX is an open, framework-neutral model representation, and ONNX Runtime provides a portable inference engine that runs on Windows, Linux, and edge devices without the original training framework. Exporting the trained model to ONNX lets the same artifact execute on-premises despite intermittent connectivity, and the team avoids re-authoring the model for each target runtime because the graph format is standardized.
- ✗
The native TensorFlow SavedModel directory
Why it's wrong here
A SavedModel is tied to the TensorFlow runtime, so serving it on-premises requires installing and maintaining that framework on every host. It does not provide runtime portability across different inference engines, which is exactly what the operations team wants to avoid. It also does not by itself reduce the model size for constrained on-premises hardware.
- ✗
A Docker image containing the full training environment
Why it's wrong here
Shipping the entire training environment as a container is heavy and couples inference to training dependencies, including GPU libraries and data tooling that are irrelevant at serving time. It also does not make the model portable across runtimes; the container still embeds one specific framework build. Image size and update friction make this a poor fit for disconnected on-premises hosts.
About these practice questions
Courseiva writes every AI0-001 question from scratch — 962 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 and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official CompTIA exam blueprint
This AI0-001 practice question is part of Courseiva's free CompTIA 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 AI0-001 exam.