AI Project Life Cycle, MLOps, and DataOps CAIPM Practice Questions (Page 6)

Questions 26–30

1. 26application · medium

   A company has separate teams for data engineering and ML engineering. The data team manages the data pipelines, and the ML team manages the model lifecycle. The company wants to implement a unified approach where data pipeline changes automatically trigger model retraining. Which integration pattern best achieves this?

   • AUse a shared database that both teams write to
   • BHave the ML team manually check for new data before retraining
   • CSchedule retraining at a fixed time regardless of data updates
   • DUse an event-driven architecture where data pipeline completion events trigger the retraining pipeline
   Reveal Solution Discussion Select an answer first
2. 27application · medium

   A startup is building its first AI product. The team consists of a data scientist, a data engineer, and a software developer. They need to define roles for the AI project life cycle. Which responsibility is MOST appropriate for the data engineer?

   • ASelecting the machine learning algorithm and tuning hyperparameters.
   • BDefining the business success criteria for the AI project.
   • CDeploying the model to production and monitoring its performance.
   • DBuilding and maintaining the data pipelines that feed the model.
   Reveal Solution Discussion Select an answer first
4. 28foundation · easy

   Which practice is a core principle of MLOps that focuses on automatically retraining models when new data becomes available?

   • AContinuous integration
   • BContinuous training
   • CContinuous monitoring
   • DContinuous delivery
   Reveal Solution Discussion Select an answer first
5. 29expert · hard

   A financial company has a credit-scoring model that is retrained monthly. The ML team wants to implement continuous integration (CI) for the model code, but the data science team is concerned that CI will slow down their experimentation because every change must pass automated tests. Which approach addresses this concern while still implementing CI?

   • AHave data scientists manually test their changes before merging
   • BRun CI tests only on the main branch after every merge
   • CSkip CI for experimental changes and only test the final model
   • DRun CI tests on a separate branch and only merge to the main branch after tests pass
   Reveal Solution Discussion Select an answer first
6. 30expert · hard

   A company is choosing tools for its MLOps stack. The team needs to support experiment tracking, model registry, and feature store. Which tool or platform provides ALL of these capabilities?

   • AVertex AI
   • BMLflow
   • CKubeflow
   • DApache Airflow
   Reveal Solution Discussion Select an answer first