SnowPro Specialty: Gen AI · 38% of the exam

Snowflake Cortex AI and LLM Functions: free practice questions

5 sample questions from our 95-question bank for this domain — answers and explanations included. These are the same scenario-based style as the real Snowflake exam.

1. In a Cortex Analyst semantic view, what is a 'dimension'?

  • A. An attribute used to group or filter by, such as region or product category✓ Correct
  • B. A raw measurable value at the row grain
  • C. A defined aggregation built on facts
  • D. A generated SQL query
Explanation

A dimension is an attribute you group or filter by. A raw measurable value is a fact, a defined aggregation is a metric, and neither is a generated SQL query.

2. A support team needs to convert incoming French customer messages into English before routing them to English-speaking agents. Which function is purpose-built for this?

  • A. TRANSLATE✓ Correct
  • B. SENTIMENT
  • C. AI_CLASSIFY
  • D. EXTRACT_ANSWER
Explanation

TRANSLATE converts text between languages, exactly this use case. SENTIMENT scores tone, AI_CLASSIFY assigns categories, and EXTRACT_ANSWER answers a question from a passage — none translate.

3. A user needs to extract a specific answer (e.g., the warranty period) from a block of product documentation text. Which Cortex function targets question-answering over a provided passage?

  • A. EXTRACT_ANSWER✓ Correct
  • B. SENTIMENT
  • C. EMBED_TEXT_768
  • D. COPY INTO
Explanation

EXTRACT_ANSWER returns the answer to a question given a source passage, ideal for pulling a specific fact from provided text. SENTIMENT scores tone, EMBED_TEXT_768 makes vectors, and COPY INTO loads data.

4. A developer wants to find 'similar' orders based on order total (a number) and considers using EMBED_TEXT_768 directly on the numeric column. Why is this the wrong tool for the job?

  • A. EMBED_TEXT_768 is designed to embed natural-language text semantics, not numeric magnitude — similarity between numbers doesn't need a text embedding function✓ Correct
  • B. EMBED_TEXT_768 requires the input to already be a VECTOR type
  • C. Numeric columns cannot be selected in Snowflake SQL
  • D. EMBED_TEXT_768 always returns a lower-dimensional vector for numeric input
Explanation

EMBED_TEXT_768 captures textual meaning, which isn't what's needed to compare numeric magnitudes — plain numeric comparison is the right tool there. It doesn't require pre-existing VECTOR input, numeric columns are selectable, and dimensionality isn't affected by input type in this way.

5. A pipeline must return the numeric sentiment of customer reviews as a score. Which Cortex function is purpose-built for this?

  • A. SENTIMENT✓ Correct
  • B. COMPLETE with no instructions
  • C. PARSE_DOCUMENT
  • D. EMBED_TEXT_1024
Explanation

SENTIMENT returns a sentiment score for input text, purpose-built for this task. PARSE_DOCUMENT extracts text from files, EMBED_TEXT_1024 produces embeddings, and a bare COMPLETE would need careful prompting to match SENTIMENT's consistency.

90 more questions in this domain

Practice the full bank with instant grading, flashcards, and a timed mock exam.

Start practicing free
Snowflake Cortex AI and LLM Functions — Free SnowPro Specialty: Gen AI Practice Questions | DataCertPrep — Certification Prep