A zoo has six weeks before a new exhibit opens and needs a visitor-FAQ helper. Which path best meets that calendar?
Select an answer to reveal the explanation.
Short Explanation
Think of a zoo with six weeks until opening day. Call a managed FM API and get the FAQ helper out. Standing up training, waiting a year for labels, or building a vision cluster misses the exhibit.
Full Explanation
Calling a managed foundation-model API best meets a zoo’s six-week opening for a visitor-FAQ helper because speed to market is a stated advantage of managed GenAI services. You can ground prompts on exhibit facts and ship without building a training plant first, which matches a fixed exhibit calendar better than a multi-month ML program.
Standing up new training infrastructure first risks missing opening day for a workload that does not require custom training to start. Waiting a year for labeled visitor questions maximizes data collection at the expense of the exhibit calendar and leaves visitors without help at opening. Building a custom vision cluster because FAQs are images misreads the problem: visitor FAQs are text questions, not a vision-training project.
Exam caveat: fast managed FAQ helpers still need factual grounding and a plan for stale exhibit details after opening week. Operational check: publish a six-week checklist covering source facts, blocked topics, sample questions, and a human escalation path for animals or safety questions the model must not invent, then re-check facts after the exhibit changes.
Speed-to-market answers on the exam usually favor managed invocation when the calendar is short and the task is FAQ-style language, not greenfield model science. Prefer grounding and review over waiting for a perfect labeled year of visitor questions that arrives after the exhibit already opened.