Tech SkillsSystem Reasoning

Contract Assistant Answers Well Question Explained

This worked AI Engineer Interview practice question tests System Reasoning. Try the answer first, then review why C is the best choice.

Short answer

The correct answer is C. Upstream of retrieval: whether footnote and table text survives OCR, structure extraction, and chunking at all..

Retrieval can only rank text that exists in the index; layout-heavy regions often die in ingestion, which makes downstream scores look 'healthy' on the text that survived. The key is to match the exact action in the prompt to the tool, command, or concept that performs that action directly. Here, C. Upstream of retrieval: whether footnote and table text survives OCR, structure extraction, and chunking at all. is the only option that satisfies the scenario without adding unrelated behavior.

The Question

A contract assistant answers well from normal paragraphs but misses obligations in footnotes and scanned tables. Retrieval scores look healthy. Where do you look first?

AThe reranker — healthy retrieval scores can still order chunks badly.
BThe prompt — the model may need explicit instructions to check footnotes.
CUpstream of retrieval: whether footnote and table text survives OCR, structure extraction, and chunking at all.Correct
DThe model — small models are known to be weak on tabular reasoning.

Why C is correct

Retrieval can only rank text that exists in the index; layout-heavy regions often die in ingestion, which makes downstream scores look 'healthy' on the text that survived. The key is to match the exact action in the prompt to the tool, command, or concept that performs that action directly. Here, C. Upstream of retrieval: whether footnote and table text survives OCR, structure extraction, and chunking at all. is the only option that satisfies the scenario without adding unrelated behavior.

Why the other options are wrong

Option A: The reranker — healthy retrieval scores can still order chunks badly.

Option A does not match the requirement in the prompt. The correct answer is C because Retrieval can only rank text that exists in the index; layout-heavy regions often die in ingestion, which makes downstream scores look 'healthy' on the text that survived.

Option B: The prompt — the model may need explicit instructions to check footnotes.

Option B does not match the requirement in the prompt. The correct answer is C because Retrieval can only rank text that exists in the index; layout-heavy regions often die in ingestion, which makes downstream scores look 'healthy' on the text that survived.

Option D: The model — small models are known to be weak on tabular reasoning.

Option D does not match the requirement in the prompt. The correct answer is C because Retrieval can only rank text that exists in the index; layout-heavy regions often die in ingestion, which makes downstream scores look 'healthy' on the text that survived.

Key idea: System Reasoning

Every question is a production situation an AI engineer interviewer can put to you: a failing endpoint, a cost spike, an agent that loops, an eval that disagrees with users. The best answer is almost always the one that measures or contains the problem before it changes the model or the prompt. If you miss a question, note which habit it tests: decompose before you fix, put the security boundary where the side effect happens, and trust an eval only after you check what it covers. The 12 open-ended questions with answer guides are at aijobprep.app/questions/ai-engineer.html.

Ready to test the full skill?

Take the free AI Engineer Interview quiz and find the exact basics to review next. No signup required.

Practice related questions

Same skill area or adjacent concept. Use these after reviewing the explanation.

Related AI Engineer Interview Practice Questions