AI Pilot Execution and Scaled Deployment CAIPM Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   An AI pilot for supply chain optimization has ended. The model met its success criteria, but the pilot team noted that data quality issues required significant manual cleanup. The program manager is preparing for scaled deployment. What should be done with the lessons learned?

   • AArchive the lessons learned in a project report and move on to deployment.
   • BDocument the data quality issues and include data pipeline improvements in the scaled deployment plan.
   • CShare the lessons learned only with the data science team, as they are technical.
   • DIgnore the data quality issues because the pilot met its success criteria.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   Which factor is most important to consider when formulating a scaled deployment strategy?

   • AThe number of pilot team members
   • BThe readiness of the organization to adopt the AI solution at scale
   • CThe color scheme of the user interface
   • DThe location of the pilot office
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   What is the primary purpose of post-deployment monitoring in an AI system?

   • ATo ensure the AI model never changes
   • BTo stop all updates to the AI system
   • CTo continuously track the AI model's performance and detect degradation
   • DTo reduce the need for user feedback
   Reveal Solution Discussion Select an answer first
5. 14expert · hard

   A logistics company successfully piloted an AI route-optimization system in one region. The pilot used a small on-premises cluster. The company now wants to scale to all regions, but some regions have strict data-residency laws that prohibit sending data to a central cloud. The company also needs low latency for real-time routing. What is the best scaled deployment strategy?

   • AMove all data to a central cloud and deploy the model globally
   • BDeploy the AI model in each region using local infrastructure that complies with data-residency laws, and use a central management plane for monitoring and model updates
   • CUse a single large on-premises cluster in one region and route all traffic there
   • DDeploy the model only in regions without data-residency restrictions
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   What is the primary purpose of scaling infrastructure for a full AI deployment?

   • ATo reduce the need for data pipelines
   • BTo eliminate the need for monitoring
   • CTo make the AI model less accurate
   • DTo ensure the system can handle increased load and data volume
   Reveal Solution Discussion Select an answer first