Translate Business Use Case Goals into a Description of the Desired Inputs and Outputs for the AI Pipeline GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A retail company wants to reduce customer churn by identifying at-risk customers each week. They have historical purchase data, customer service interaction logs, and demographic data. The business team expects a list of customer IDs with a churn risk score. What is the most appropriate measurable objective for this AI pipeline?

   • APredict the exact date each customer will churn.
   • BProvide a dashboard showing overall churn rate trends.
   • CSegment customers into high, medium, and low value groups based on purchase history.
   • DGenerate a weekly list of customer IDs with a churn risk score between 0 and 1, where a higher score indicates higher risk.
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   An AI pipeline is being designed to classify support tickets by urgency. Which input specification is most complete?

   • AA CSV file containing ticket data.
   • BText data that has been cleaned and tokenized.
   • CAll support tickets.
   • DTicket text from the CRM system in JSON format, with preprocessing to remove HTML tags and tokenize the text.
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   A bank wants to use AI to approve loan applications. Which output specification best meets the business goal?

   • AA binary decision ('approved' or 'denied') for each application, along with a confidence score.
   • BA probability of default for each applicant, without any decision threshold.
   • CA report summarizing the average loan amount.
   • DA list of all applicants sorted by credit score.
   Reveal Solution Discussion Select an answer first
5. 9expert · hard

   A news organization wants to use AI to summarize articles for a mobile app. The goal is to increase user engagement. They have a large corpus of articles and user reading data. The app team needs summaries that are concise (under 50 words) and accurate. However, the organization has a policy that all summaries must be fact-checked by editors before publication. What output specification best balances the business goal and the editorial policy?

   • APublish the AI-generated summary directly without human review.
   • BGenerate summaries only for articles that the model is highly confident about, and publish those directly.
   • CGenerate a summary and automatically send it to the app without editor review.
   • DGenerate a draft summary for each article, with a confidence score, for editors to review and approve before publication.
   Reveal Solution Discussion Select an answer first
6. 10expert · hard

   A nonprofit organization wants to use AI to identify donors likely to donate again. The goal is to increase fundraising efficiency. They have donation history, event attendance, and email engagement data. The fundraising team wants a list of top 100 donors to contact each month. However, the organization has a strict privacy policy that prohibits using certain demographic data. What is the most appropriate measurable objective?

   • AGenerate a monthly report of donation trends by donor segment.
   • BGenerate a list of all donors with a donation probability score, regardless of data sources.
   • CGenerate a list of donors who have not donated in the past year.
   • DGenerate a monthly list of the top 100 donors by predicted likelihood to donate, using only permitted data sources.
   Reveal Solution Discussion Select an answer first