AI0-001 AI Infrastructure and Technologies Practice Question
A financial institution is deploying a real-time anomaly detection model on a Kubernetes cluster. The model must process streaming transactions with low latency and scale horizontally during peak hours. The team wants to use a serving solution that integrates natively with Kubernetes and supports autoscaling based on request concurrency. Which solution best meets these requirements?
⚠ Common exam trap
The trap here is assuming that deploying any model server on Kubernetes automatically provides request-concurrency autoscaling, when it often requires additional components.
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
✓
KServe with Knative autoscaling
KServe with Knative autoscaling provides native Kubernetes integration and request-concurrency-based scaling, which is essential for handling unpredictable transaction volumes with low latency. The other options either lack autoscaling, require manual configuration, or are not optimized for production inference.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
KServe with Knative autoscaling
Why this is correct
KServe integrates with Kubernetes and uses Knative to provide request-based autoscaling, including scale-to-zero and concurrency metrics. It supports low-latency inference and horizontal scaling, directly matching the streaming transaction requirements. This makes it the most appropriate choice.
- ✗
Standalone TensorFlow Serving deployed as a Kubernetes Deployment
Why it's wrong here
TensorFlow Serving can run on Kubernetes, but it lacks native request-concurrency-based autoscaling. The team would need to configure Horizontal Pod Autoscaler manually using CPU or custom metrics, which adds complexity and may not react quickly enough. It does not provide the integrated autoscaling described.
- ✗
NVIDIA Triton Inference Server with a fixed number of replicas
Why it's wrong here
Triton offers high-performance inference but does not natively autoscale based on request concurrency. A fixed replica count cannot handle peak-hour spikes efficiently, leading to either over-provisioning or latency degradation. Additional tooling would be required to achieve the desired elasticity.
- ✗
A custom Flask API wrapping the model, deployed with a Kubernetes Service
Why it's wrong here
A custom Flask API may work for simple cases but lacks optimized inference serving, batching, and built-in autoscaling. It would require significant engineering to match the performance and scalability of a dedicated serving platform. This approach increases operational risk and is not suitable for low-latency, high-scale streaming.
About these practice questions
This AI0-001 question is part of Courseiva's 962-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam 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.