nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #251

You work at a mobile gaming startup that creates online multiplayer games. Recently, your company observed an increase in players cheating in the games, leading to a loss of revenue and a poor user…

The correct answer is D. Import the model into Vertex AI Model Registry. Create a Vertex AI endpoint that hosts the model. To provide immediate, low-latency classifications for a trained model in production, the model should be imported into Vertex AI Model Registry and deployed to a Vertex AI endpoint for online prediction.

Submitted by klara.se· Apr 18, 2026ML pipeline operationalization

Question

You work at a mobile gaming startup that creates online multiplayer games. Recently, your company observed an increase in players cheating in the games, leading to a loss of revenue and a poor user experience You built a binary classification model to determine whether a player cheated after a completed game session, and then send a message to other downstream systems to ban the player that cheated. Your model has performed well during testing, and you now need to deploy the model to production. You want your serving solution to provide immediate classifications after a completed game session to avoid further loss of revenue. What should you do?

Options

  • AImport the model into Vertex AI Model Registry. Use the Vertex Batch Prediction service to run
  • BSave the model files in a Cloud Storage bucket. Create a Cloud Function to read the model files
  • CSave the model files in a VM. Load the model files each time there is a prediction request, and
  • DImport the model into Vertex AI Model Registry. Create a Vertex AI endpoint that hosts the model,

How the community answered

(18 responses)
  • A
    6% (1)
  • B
    6% (1)
  • C
    11% (2)
  • D
    78% (14)

Why each option

To provide immediate, low-latency classifications for a trained model in production, the model should be imported into Vertex AI Model Registry and deployed to a Vertex AI endpoint for online prediction.

AImport the model into Vertex AI Model Registry. Use the Vertex Batch Prediction service to run

Vertex Batch Prediction is designed for high-throughput, asynchronous predictions on large datasets, not for immediate, low-latency, real-time classifications.

BSave the model files in a Cloud Storage bucket. Create a Cloud Function to read the model files

While Cloud Functions can serve models, directly loading model files from Cloud Storage for each request can introduce higher latency and isn't as optimized or scalable for real-time serving as a dedicated Vertex AI endpoint.

CSave the model files in a VM. Load the model files each time there is a prediction request, and

Saving model files in a VM and loading them per request is a manual, unmanaged approach that lacks scalability, high availability, and the managed features of Vertex AI for production model serving.

DImport the model into Vertex AI Model Registry. Create a Vertex AI endpoint that hosts the model,Correct

Importing the model into Vertex AI Model Registry centralizes model management. Creating a Vertex AI endpoint to host the model enables online predictions, which are suitable for immediate, low-latency classifications required after a completed game session. This setup allows for real-time inference and integrates well with downstream systems for actions like banning players.

Concept tested: Vertex AI online prediction for low-latency inference

Source: https://cloud.google.com/vertex-ai/docs/predictions/online-predictions

Topics

#Model deployment#Online prediction#Vertex AI Endpoints#Low latency

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice