AI Project Life Cycle, MLOps, and DataOps CAIPM Practice Questions (Page 8)

Questions 36–40

1. 36expert · hard

   A company is building a real-time fraud detection system. The data pipeline and model are updated frequently. The team wants to ensure that the model always uses the latest features and that deployments are seamless. Which architecture BEST supports this?

   • AA separate data pipeline and model serving system with no shared feature store.
   • BA monolithic application that embeds the model and data processing.
   • CA batch data pipeline that updates features daily and manual model deployment.
   • DA feature store with online and offline consistency, and a CI/CD pipeline for model deployment.
   Reveal Solution Discussion Select an answer first
2. 37foundation · easy

   In an AI project team, which role is typically responsible for operationalizing machine learning models, including deploying them to production and monitoring their performance?

   • ABusiness analyst
   • BData engineer
   • CData scientist
   • DML engineer
   Reveal Solution Discussion Select an answer first
4. 38foundation · easy

   What is a common challenge when implementing MLOps in an organization, and what is a best practice to overcome it?

   • AChallenge: high cost of cloud computing; Best practice: avoid using cloud services entirely
   • BChallenge: slow model training; Best practice: increase the size of the training dataset
   • CChallenge: difficulty in reproducing models; Best practice: use containerization and versioning for code, data, and models
   • DChallenge: lack of collaboration between teams; Best practice: implement version control for code and data
   Reveal Solution Discussion Select an answer first
5. 39application · medium

   A logistics company is building an AI system to optimize delivery routes. The team needs to manage data pipeline orchestration, model versioning, and automated testing of both code and data. Which combination of tools would BEST support these needs?

   • ADocker for orchestration, GitHub for data versioning, and cron jobs for CI/CD.
   • BApache Airflow for orchestration, DVC for data versioning, and Jenkins for CI/CD.
   • CApache Kafka for orchestration, Git for data versioning, and a spreadsheet for tracking.
   • DTableau for orchestration, MLflow for data versioning, and manual deployment.
   Reveal Solution Discussion Select an answer first
6. 40application · medium

   An online media company updates its recommendation model frequently. The ML team wants to automate the pipeline from code commit to model deployment in production. Which MLOps capability is ESSENTIAL to achieve this?

   • AA data lake to store all training data.
   • BA feature store to manage features.
   • CA dashboard to visualize model performance.
   • DA CI/CD pipeline that builds, tests, and deploys the model automatically.
   Reveal Solution Discussion Select an answer first