The latest changes and updates from the administration for this exam.
Latest Update: Jun 08 2026
All questions are working fine.
You're a data engineer at a large e-commerce company, and you've recently trained a product recommendation machine learning model using Google Cloud AI Platform. The model performs well during testing, and you're ready to deploy it. However, the number of users and product diversity of your company requires this model to be highly scalable and serve low-latency predictions. Which deployment method would best meet these requirements?
You are tasked with migrating an existing on-premises columnar data warehouse to Google Cloud. The existing warehouse contains 200 TB of data. The organization requires minimal downtime and wants to implement a solution that will enable delta updates in the future to maintain synchronization with the on-premise system while they complete their migration to the cloud. What is the best approach to meet these requirements?
A retail company wants to build a real-time customer analytics system to track customer behavior, including product purchases, website visits, and customer demographics. The system should be scalable, easy to maintain, and able to handle large amounts of data. What is the most appropriate solution for this scenario from the following?
You are a data engineer for a large healthcare organization that generates significant amounts of data from wearable devices. The organization intends to use this data for real-time health monitoring and anomaly detection. Because of privacy concerns, and the need for real-time predictions, the decision is to perform the data processing and machine learning inference on the edge devices themselves. What infrastructure would you recommend for this task?
You're a data engineer at a large enterprise and your team has built an ML model that predicts sales based on various factors. The model has been deployed on AI Platform Prediction and has been performing well. However, due to recent changes in business trends, the model's predictions have started to deviate significantly. You've decided to retrain the model every week with the latest data, which is stored in BigQuery. How would you set up this retraining and redeployment process?
A healthcare company wants to store and process large amounts of patient medical records, including electronic health records (EHRs), lab results, and imaging studies. The solution should be secure, scalable, and compliant with industry regulations such as HIPAA. What is the most appropriate solution for this scenario from the following?
You are building a data processing pipeline to analyze large amounts of event data generated by a mobile application. The event data is stored in Google Cloud Storage and needs to be processed in real-time for use in a recommendation engine. What steps would you take to build this pipeline?
Your company operates in multiple countries and has a strict data residency requirement to keep data within the borders of the country of origin. It also wants the flexibility to move to different cloud providers or return to on-premises if needed. The requirement is to design a data storage and processing system that meets these requirements. What should you recommend?
You are a data engineer at a large corporation and you have been tasked with developing an end-to-end Machine Learning pipeline in Google Cloud. The company has a large set of training data stored in BigQuery and uses AutoML Tables to train its models. As the models update frequently, you need to automate the pipeline as much as possible. Which approach should you use for this task?
A social media company wants to store user data for their social media platform using NoSQL document database. The data should be highly scalable and available, and it should allow for fast querying and retrieval of data by multiple users simultaneously. What solution would you recommend for this use case to store the user data while ensuring high scalability, availability, and fast querying and retrieval?
