Manage Model Lifecycle and Governance GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 4)

Questions 16–20

1. 16application · medium

   A company uses a Vector Search index to power a retrieval-augmented generation (RAG) application. They update the underlying source data nightly and want the index to reflect these changes automatically as part of their deployment pipeline. Which CI/CD practice should they implement?

   • AUse the MLflow model registry to store the index and update it via model versioning.
   • BManually recreate the index from the source data every time the data changes.
   • CEmbed the index update in the application code and redeploy the application each time.
   • DTrigger a pipeline job that runs a notebook to sync the index from the source data after each data update.
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A prompt has been in production for several months but is now causing incorrect responses due to changes in the underlying data. The team decides to retire the prompt and replace it with a new one. What should they do to manage this transition properly?

   • AKeep the old prompt in production and add the new prompt as an alternative, letting users choose.
   • BDeprecate the old prompt in the prompt registry, document the reason, and deploy the new prompt through the standard review process.
   • CUpdate the old prompt in place without versioning or review.
   • DDelete the old prompt immediately and replace it with the new one in the same commit.
   Reveal Solution Discussion Select an answer first
4. 18foundation · easy

   Which of the following is a key characteristic of a persistent datastore used for agent memory?

   • AIt stores data only for the duration of a single request
   • BIt is only accessible from the compute node that created it
   • CIt is automatically cleared when the agent finishes its task
   • DIt retains data across sessions and can be queried for retrieval
   Reveal Solution Discussion Select an answer first
5. 19expert · hard

   A team uses a Git repository to manage prompts. They want to promote a prompt from development to production, but they need to ensure that the prompt is reviewed by a senior engineer and that the deployment is automated. Which CI/CD practice should they implement?

   • AUse a pull request for review and a CI/CD pipeline that deploys the prompt to production after the pull request is merged.
   • BHave the senior engineer manually copy the prompt to production after reviewing it.
   • CUse a shared notebook and have the senior engineer approve changes in the notebook.
   • DDeploy the prompt to production automatically without review to speed up the process.
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A prompt has been deprecated but is still referenced by some legacy applications. The team wants to ensure that the deprecated prompt is not used by new applications. What should they do?

   • AChange the prompt's content to a placeholder to discourage use.
   • BKeep the prompt unchanged and do nothing.
   • CDelete the prompt immediately to prevent any further use.
   • DMark the prompt as deprecated in the prompt registry and communicate the deprecation to the development teams.
   Reveal Solution Discussion Select an answer first