Which service returns identical answers to HR policy questions from PDFs every time?
You are designing a solution that will answer questions about human resources (HR) policies stored in the PDF format. You need to ensure that the identical answer to a specific question is returned every time. The solution must minimize development effort. Which service should you include in the solution?
Community Insight
The phrase "identical answer every time" signals extractive, knowledge-base question answering, and the trap is reaching for Azure OpenAI, whose generative responses are non-deterministic by design.
This question tests whether you can pick the Foundry Tools service that gives deterministic, repeatable answers over HR policy PDFs with minimal build effort. The answer is Azure Language in Foundry Tools, because its custom question answering knowledge base returns fixed extractive answers rather than freshly generated wording.
Most candidates pick Azure OpenAI because it feels like the modern way to answer questions over documents, but generative completions reword the response on every call and therefore cannot guarantee the identical answer each time.
Community Discussion (2 comments)
Comments & Corrections
No comments yet — spotted an error or have a note? Share it below.
Expert Analysis
Why the Answer Is Correct
Azure Language in Foundry Tools includes the question answering capability (formerly QnA Maker), which ingests HR policy PDFs into a curated knowledge base and returns the matched answer or source passage as-is. Because the service performs retrieval and matching against stored answer pairs rather than generating new text, the same question yields the same answer on every call. Importing PDFs and DOCX files as sources means very little custom code, satisfying the minimize-development-effort requirement. You can further harden determinism with alternate questions, metadata filters, and confidence thresholds.Why the Other Options Are Wrong
Azure OpenAI generates responses token by token, so temperature, sampling, and prompt context can change the phrasing of otherwise correct HR policy answers. Azure Machine Learning gives you full control but requires designing, training, and hosting your own retrieval and answering pipeline, which is the opposite of minimizing effort. Azure Document Intelligence in Foundry Tools is a strong extractor of text, tables, and layout from PDFs, yet it has no knowledge base or question-to-answer matching layer, so it can feed a QA pipeline but cannot be the answering service itself. Only option B combines deterministic answers with a managed knowledge base over PDFs.Community Comment Notes
Both commenters land on B and their reasoning matches the exam doctrine. Srikantha frames question answering as "matching user queries to static, predefined answers," which is exactly why repeated questions return identical output. Az301301X isolates the decisive requirement and concludes it "points to Exact Question Answering, not generative AI." The agreement here is useful corroboration, but the verdict rests on the deterministic-answer wording of the scenario rather than on vote counts.Official Reference
Exam Strategy
When a stem demands the same answer every time, treat it as a non-generative requirement and eliminate any model that composes fresh text. Look for the service that stores answers in a knowledge base and returns passages verbatim, and remember that Document Intelligence only extracts, it does not answer.
Frequently Asked Questions
Why is Azure OpenAI wrong when it can also read the HR policy PDFs?
Generative deployments sample new wording on every call, so the same HR question can produce different phrasing, which breaks the identical-answer requirement.
What role could Document Intelligence in Foundry Tools play alongside question answering?
It can preprocess scanned or complex HR PDFs into clean text, but it has no knowledge base or answer-matching layer, so it cannot guarantee repeatable answers alone.
Related Analysis
Practice All AI-103 Questions
Access 88 questions with complete answers and detailed explanations.
View Full AI-103 Practice Test →