Independent: not affiliated with AWS. Free to try; the study bundle adds a study guide and two practice exams.
5 original questions, none of them in the practice exams.
Pick an answer. Show answer explains every option and links its AWS source.
5 questions / Explanations link AWS sources / Last updated Oct 26, 2026
Question 1
Multiple choice
Domain 1 / Task 1.2 Practical use cases for AI
A dental practice wants a bot on its website and phone line that lets patients schedule, move or cancel a visit. The bot must ask for the date, the time and the patient's name, then pass them to the practice's booking system. Which AWS service is built for this?
Show answerHide answer
Answer: C
Why
Amazon Lex builds chatbots and voice bots from intents, slots and sample utterances. Scheduling, moving and cancelling a visit are intents, the date, time and name are slots, and a Lambda function can hand them to the booking system.
Why not the others
ATurns text into speech. It can voice a reply but does not run the conversation.
BTurns speech into text. It does not collect slots or decide what to ask next.
DMakes recommendations from interaction data, not conversations.
An internal tool sends each of yesterday's support tickets to an Amazon Bedrock model for a summary, one API call per ticket. The team will accept slower responses for a lower price per token, and the tool cannot be rebuilt to read and write files in Amazon S3. Which option fits?
Show answerHide answer
Answer: A
Why
The Flex tier is on-demand inference at a 50% discount to Standard pricing, for work that tolerates slower processing. The tool keeps making its usual API calls.
Why not the others
BCosts 75% more so that requests are served first. The team wants the opposite trade.
CAlso discounted, but it runs asynchronous jobs that read and write files in S3, which the tool cannot do.
DBills per model unit per hour for guaranteed throughput, which nothing here calls for.
Domain 3 / Task 3.1 Design considerations for FM applications
A manufacturer is building an Amazon Bedrock Knowledge Base over its repair manuals. Technicians look up exact part codes such as HX-4410, and they also ask plain-language questions such as 'Why does the pump overheat?' Which vector store fits best?
Show answerHide answer
Answer: B
Why
Exact codes need keyword matching, and plain-language questions need vector similarity. OpenSearch Service runs both in one hybrid query, and it is the quick-create vector store for Knowledge Bases.
Why not the others
AChosen for the lowest storage cost on very large sets that are queried infrequently, not for keyword matching.
CFits questions about how records are connected, which these are not.
DNot a Knowledge Bases vector store. In a RAG design it holds chat history or metadata.
Domain 4 / Task 4.1 Developing responsible AI systems
A city launches an assistant that answers residents' questions about permits. In a review, staff find they have no way to pause it, override one of its replies or roll back a bad update. Which of AWS's core dimensions of responsible AI is this design missing?
Show answerHide answer
Answer: C
Why
Controllability means having mechanisms to monitor and steer an AI system's behavior. With no way to pause, override or roll back, staff cannot steer this one.
Why not the others
AAbout telling people they are using AI and what it cannot do.
BAbout the system's impacts on different groups of people.
DAbout correct outputs, even with unexpected or adversarial input.
An HR assistant built on an Amazon Bedrock Knowledge Base answers staff questions. In testing, an employee asked about a colleague and got back the colleague's home address and national ID number, taken from a file only HR managers may open. Which two measures most directly prevent this? Choose two.
Show answerHide answer
Answers: A and C
Why
Data leakage is stopped at two points: retrieval returns only what the person asking may see, and a sensitive information filter masks or blocks PII in what the model sends back.
Why not the others
BLogging keeps a record of the leak for audit. It does not prevent it.
DTemperature changes how random the output is, not which documents reach the model.
EArtifact provides AWS's own compliance reports. It does not control your data.
The study bundle has two practice exams with 130 original questions weighted to the five exam domains, an answer key and an explanation for every answer, plus the 54-page study guide. Practice exam explanations are shorter than the ones on this page: a short paragraph per question, with no source links.
Study bundle
$19
130 practice questions / Study guide / Answer keys
Every bundle includes 6 months of access: download the PDFs, and any new editions, until 6 months after your purchase date. Your receipt has a license key. Enter it at overqual.com/download for the three PDFs, stamped with your email and order number. They are yours to keep.