Task 2.1: Implement Agentic AI Solutions and Tool Integrations. AIP-C01 Practice Questions (Page 7)

Questions 31–35

1. 31foundation · easy

   When would you choose Amazon ECS over AWS Lambda for implementing an MCP server?

   • AWhen you need to run a short-lived function
   • BWhen the tool requires more resources or stateful processing
   • CWhen you want to minimize operational overhead
   • DWhen you want to avoid managing servers
   Reveal Solution Discussion Select an answer first
2. 32foundation · easy

   What is the goal of human augmentation patterns in AI systems?

   • ATo combine human expertise with FM capabilities to achieve better outcomes than either alone
   • BTo automatically generate training data
   • CTo reduce the need for data storage
   • DTo replace human workers entirely with AI
   Reveal Solution Discussion Select an answer first
4. 33foundation · easy

   Which of the following is a benefit of using MCP client libraries?

   • AAutomatic scaling of the MCP server
   • BGuaranteed zero-latency access to tools
   • CElimination of all security risks
   • DReduced development time by abstracting protocol details
   Reveal Solution Discussion Select an answer first
5. 34expert · hard

   A media company is building a system that generates product descriptions. They use two different foundation models: one is fast and cheap but sometimes produces inaccurate claims, and the other is slower and more expensive but highly accurate. They want to combine the outputs to improve overall quality while keeping costs reasonable. They are considering a custom aggregation logic that uses the fast model's output as a draft and the slow model's output to verify and correct it. What is the most important consideration when implementing this aggregation logic?

   • AAlways use the slow model's output as the final result, regardless of what the fast model produces.
   • BRandomly select one of the two models' outputs to use as the final result to avoid bias.
   • CUse the fast model's output as the final result and only log the slow model's output for auditing purposes.
   • DEnsure the aggregation logic can handle cases where the two models produce conflicting outputs and define a clear rule for resolving the conflict.
   Reveal Solution Discussion Select an answer first
6. 35expert · hard

   A company is building an AI system that performs two tasks: summarizing legal documents and generating marketing copy. They have two foundation models: Model A is excellent at summarization but poor at creative writing, and Model B is excellent at creative writing but poor at summarization. The team wants to use a single model selection framework to choose the best model for each task. What is the most important factor to consider?

   • AThe framework should use an ensemble of both models for every task, combining their outputs to get the best of both.
   • BThe framework should route tasks to the model that is specialized for that task, using Model A for summarization and Model B for marketing copy.
   • CThe framework should always use Model A because it is better at summarization, which is a more important task.
   • DThe framework should randomly select a model for each task to avoid bias.
   Reveal Solution Discussion Select an answer first