AI Strategy and Adoption Roadmap Design CAIPM Practice Questions (Page 7)

Questions 31–35

1. 31application · medium

   A logistics company has deployed an AI system that predicts delivery delays. The operations team wants to know if the system is helping them meet their service-level agreement (SLA) of 95% on-time delivery. Which KPI is most appropriate for measuring the AI system's contribution?

   • ANumber of model versions deployed
   • BPercentage of on-time deliveries before and after AI deployment
   • CModel prediction speed in milliseconds
   • DNumber of data scientists assigned to the project
   Reveal Solution Discussion Select an answer first
2. 32expert · hard

   A mid-sized company has a $500,000 annual AI budget. The roadmap has three phases: Phase 1 (foundation, $200k), Phase 2 (pilot, $150k), and Phase 3 (scale, $150k). The CFO wants to cut the budget by 20% but still achieve a working pilot. The data team says Phase 1 is essential for data quality. Which budget adjustment is most appropriate?

   • ADistribute the 20% cut evenly across all three phases
   • BReduce Phase 1 to $100k and increase Phase 2 to $200k to accelerate the pilot
   • CReduce Phase 3 to $50k and keep Phase 1 and Phase 2 at their original amounts
   • DCut Phase 2 to $50k and keep Phase 1 at $200k
   Reveal Solution Discussion Select an answer first
4. 33foundation · easy

   What is the purpose of establishing feedback mechanisms in an AI roadmap?

   • ATo gather input for continuous improvement
   • BTo finalize the roadmap permanently
   • CTo reduce stakeholder involvement
   • DTo eliminate the need for monitoring
   Reveal Solution Discussion Select an answer first
5. 34foundation · easy

   Which risk mitigation strategy is commonly used to address potential bias in AI models?

   • AConducting regular bias audits
   • BRemoving all human review
   • CUsing only one data source
   • DIgnoring bias until deployment
   Reveal Solution Discussion Select an answer first
6. 35application · medium

   A hospital has two AI use cases: (1) predicting patient readmission risk and (2) automating appointment scheduling. The readmission model requires integrating data from three legacy systems and has a high potential to reduce costs. The scheduling automation is simple, uses existing data, and has moderate cost savings. The hospital has limited IT resources. Which use case should be prioritized first?

   • AReadmission prediction, because it has higher potential cost reduction
   • BRun both projects in parallel to maximize value
   • CAppointment scheduling, because it is feasible with existing data and resources
   • DOutsource the readmission model to a vendor to avoid internal resource constraints
   Reveal Solution Discussion Select an answer first