Google Professional Machine Learning Engineer Question 299
Select 2Google Cloud PlatformYou are working as a Machine Learning Engineer for a retail company that wants to implement a recommendation system. The system needs to provide real-time recommendations to users browsing the website and generate daily batch reports for analyzing overall user engagement trends. Which combination of Google Cloud services would best meet these requirements?
- A
Use Vertex AI to serve the real-time recommendation model for online inference and Dataflow to generate daily batch reports.
- B
Use BigQuery ML to train the recommendation model and schedule SQL queries for daily batch reports.
- C
Use Dataproc to train the recommendation model and deploy it for real-time inference.
- D
Use Vertex AI for training and deploying the recommendation model for real-time inference and BigQuery for generating daily batch reports.
- E
Use Dataflow for both real-time inference and batch reporting by configuring separate pipelines.
Show answer and explanation
Correct answers: A, D
Explanation
The problem requires a solution for both real-time inference and batch reporting. Vertex AI is a managed service designed for deploying machine learning models for online inference, making it suitable for real-time recommendations. For batch reporting, using Dataflow or BigQuery to process and analyze data is appropriate. Option 1 leverages Vertex AI for real-time inference and Dataflow for batch reporting, while Option 4 uses Vertex AI for real-time inference and BigQuery for batch reporting. Both are valid solutions that align with Google Cloud best practices.
- A. Correct.
This option is correct. Vertex AI is designed for deploying machine learning models for online (real-time) inference, and Dataflow is a suitable choice for processing data in batch mode to generate daily reports.
- B. Incorrect.
This option is partially correct but not optimal for real-time inference. BigQuery ML is better suited for training models and performing batch predictions within BigQuery rather than handling real-time inference.
- C. Incorrect.
This option is incorrect. While Dataproc is useful for distributed data processing and training, it is not typically used for serving real-time inference.
- D. Correct.
This option is correct. Vertex AI can handle real-time inference effectively, and BigQuery is excellent for running scheduled queries to generate daily batch reports.
- E. Incorrect.
This option is incorrect. While Dataflow is a robust tool for batch and streaming data pipelines, it is not designed to serve real-time ML models for inference. Separate services like Vertex AI are preferred for real-time ML serving.